ARTICLES

 

The loyalty of the literary reviser: Author, source text, target text or reader?

 

 

Ilse Feinauer^I; Amanda Lourens^II

^IDepartment of Afrikaans and Dutch, Stellenbosch University, South Africa E-mail: aef@sun.ac.za
^IIDepartment of Afrikaans and Dutch, Stellenbosch University, South Africa E-mail: alourens@sun.ac.za

 

 

---

ABSTRACT

The processes of revision and translation, according to Mossop (2010:112-113), can address the problem of conflicting interests, goals and needs by taking different approaches. Translation, he suggests, should seek to achieve a balance between loyalty to the source text author and to the target text readers, whereas revision should serve the interests of the future readers of the text. As a result, revising activities will steer away from a linguistic or a text-based approach in order to prioritise the needs of the reader. The question, however, is whether revisers in literary translation processes do follow the suggested approach and prioritise the needs of the target readers. An empirical analysis of the metatextual discourse among the agents involved in three different literary translation processes seeks to answer this question. During the course of this analysis, a second question presents itself, namely whether self-revision and other-revision (Mossop 2010:167, 174) should be discerned as two distinct types of revision or whether this distinction could be refined. The results of the subsequent analysis give rise to the presumption that such a binary projection of the act of revision can be contested. It is instead suggested that revision can be plotted on a continuum, with self-revision by the translator as the one end and revision by others as the other end. The analysis of three Afrikaans novels translated into English by the same award-winning translator suggests that self-revision by the translator may find itself moving away from true self-revision (a process that is not influenced by feedback from agents other than the translator) initially to a second phase of self-revision that is shaped by revision by others (e.g. the author, reviser and editor).

Keywords: agent, archival material, literary translation, loyalty, revision

---

 

 

1. Introduction

This paper presents a study of the process of the professional revisions performed on three works of fiction translated from Afrikaans into English in a fairly large book production company in South Africa that employs freelance translators, revisers¹ and editors.²

The study aims to shed some light on the focus of the loyalty³ of the literary reviser - does his or her loyalty during the revision of the draft lie with the source text author, with the source text, with the target text or with the target text reader? Such an investigation would address the question of how revisers seek to solve the problem of what Mossop (2010:112) calls a conflict of interests, goals and needs (thus implying different loyalties). Do the revisers perhaps favour the future readers of the text, while the translator focuses on accuracy, correct and idiomatic language and the client's (publisher's or author's) instructions, as suggested by Mossop (2010:112-113), as one way of dealing with the problem of conflicting interests?

An empirical analysis of the documentation exchanged among the agents involved in three different literary translation projects seeks to answer this question. Specifically, the revisions of three Afrikaans novels that had been translated into English by the same award-winning translator were analysed in the light of Buzelin's (2007, 2011 in Scocchera 2016) view that translation and revision are mainly human activities which are best investigated by eliciting data directly from its agents and in real-life contexts.

 

2. Literature review and problem statement

An important aspect of the text-production process addressed by Mossop (2010:21, 112) is the fact that the needs of the different parties involved in the process might be in conflict. In a real-life situation, an author might insist that foreign terms and references be retained in a translated text, while the translator and the reviser might be of the opinion that these might alienate the target readers. For Mossop, the solution lies in a balancing of interests of the respective parties making an input into the translation process. These parties, for instance, may be the author of the source text or the publisher and the revisers of the particular translation.

Mossop (2010:17) makes it clear that revision relies on two functions, namely that of the so-called gatekeeper and the language therapist. The gatekeeper function aims at correcting the text and bringing it into line with pre-existing rules (e.g. the rules of the language or the house style of the publisher). The language therapist effects those changes that can be deemed not absolutely necessary; rather, they are the changes that assist the reader in his or her journey through the text. Words that seem too foreign might be substituted with more indigenous terms, and complex sentences might be simplified. These changes are not dictated by any specific manual, as might happen during gatekeeping, but the editor or reviser will anticipate the needs of the readers and adjust the text in order to be more accessible for them.

The problem of conflicting interests, according to Mossop (2010:112-113), can be solved⁴ by the translator and reviser adopting different stances. The translator will try to find a balance between the author and the reader or between a retrospective and prospective orientation (thereby addressing what Mossop (2010:112) calls the "central problem of translation"). While the translator should try to find a balance between these two opposing demands, Mossop suggests that the reviser should adopt a different approach. Instead of simply acting as a second translator, he or she shifts to another position in the social network that is being constructed, namely one where he or she leans towards the needs of the readers and tries to see the text from their point of view. Mossop (2010:113) states, "[T]he reviser tries to meet the needs of readers always, and the needs of others if possible" (although who these "others" are is never clarified). By shifting his or her loyalty to the readers, Mossop contends, a balance between the interests of author and reader can be achieved. Eventually, Mossop's assumed model in which the translator is traditionally more focused on gatekeeper tasks, while the reviser takes on the role of language therapist, seems to propose a certain binarity regarding the tasks of translation and revision.

This study was prompted by the question of whether such a finely coordinated balance is indeed achieved during the translation and revision of actual texts. Do actual translation and revision processes follow a route that is shaped by a consciously decided sequence of loyalties, as proposed by Mossop (2010)? Or is it possible that such a sequence is not recognised in actual processes, as might be suggested by Mossop's (2014) omission of his previous proposition?

The current study therefore set out to investigate the loyalties of the different agents in three different translation production processes that were undertaken and completed by a South African publishing house. In all the cases, a work of fiction was translated and revised through the participation of various agents. In this study, these agents are categorised according to the terms used by the publishing house, namely 'author', 'translator', 'reviser', 'editor', 'compiler' and 'proofreader'. However, all activities of these agents dealing with the draft translation in order to produce a better quality target text are referred to as 'revision'.

The nature of the texts under investigation has, however, given rise to questions about the influence of genre on the balancing act described by Mossop. In an informative text such as a brochure on a health issue for a mass readership, it makes perfect sense that the interests of the client and reader are balanced. But what about the case of literary texts, in which the author sometimes deliberately strives to perplex his or her readers? Will the needs of the readers then be interpreted as having the perplexing sections explained to them? Or are their needs specifically to be challenged by the text, meaning that explicitation by for example the reviser might actually not take place in favour of their needs?

 

3. Methodology

Following the approach adopted by Munday (2012) in his case studies of three literary translation and/or revision processes, this study utilised archival material in order to study three different sets of agents involved in the translation process. (According to Munday (2012:104), archival documents have been underutilised in translation studies, even though they hold the possibility of providing detailed retrospective insight into the decision-making processes involved in translation and revision.) Specifically, the e-mail correspondence among the agents involved in the translation and revision of three Afrikaans works of fiction⁵ was studied in order to answer the question regarding the loyalty of the different agents. The only constant factor across these processes is the translator, who was responsible for all three translations.

The first text studied is the award-winning Afrikaans novel Niggie by Ingrid Winterbach (2002), translated into English by Elsa Silke as To hell with Cronjé (2007, 2010). Set against the backdrop of the South African War, the novel deals with the traumatic and sometimes bizarre experiences of two scientists towards the end of the war. The source text was awarded the prestigious Hertzog Prize for literature in 2004 and occupies a prominent position as a serious literary text in the Afrikaans literary polysystem. Equally, Winterbach can be seen as occupying a significant position in the highly canonised strata of the Afrikaans system, with both her and the source text being rich in symbolic capital (Spies 2013:191-192).

The second text is the award-winning Afrikaans youth novel Vaselinetjie (2004) by Anoeschka von Meck, translated by Elsa Silke as My name is Vaselinetjie (2009). It tells the coming-of-age story of an abandoned white baby girl who was raised by a coloured couple and, in a heartbreaking turn of events, was taken away by child welfare services and sent to a state orphanage at the age of 11. Vaselinetjie was awarded the prestigious MER Prize for youth literature as well as the Jan Rabie/Rapport Prize in 2005. The source text is canonised as a youth novel that has been prescribed at high school level but is also seen as a 'crossover' book that both teenagers and adults can relate to (Spies 2013:193-194).

Lastly, the article examines the revision processes of the volume of short stories published in English as In bushveld and desert: A game ranger's life (Bakkes 2008). For this volume, a number of stories from a novel and three volumes of short stories by Christiaan Bakkes (2004, 2006, 2007), who is well known as a seasoned traveller and game ranger in Africa, were selected by the publisher and translated by Elsa Silke. Bakkes, not having won any literary prizes, does not enjoy the same status in the Afrikaans literary system as, for example, Winterbach. The target readers of In bushveld and desert are people who enjoy well-written stories about Africa and nature, and especially tourists in Southern Africa. (Spies 2013:195-196).

The following sets of comments or notes (in the form of e-mail correspondence) by the various agents all working as revisers⁶ on the drafts were obtained for the three texts that were translated:

 

 

 

 

 

 

It is not clear whether the different agents involved in the three translations were provided with a brief that set out their roles and responsibilities.

Each individual comment/note was tagged (e.g. M1), coded according to six categories representing the six possible loyalties of the agents that emerged from the comments. Totals for the six categories were calculated - for the individual texts and for all three texts collectively.

These categories are the (target) language, the source text, the target text, the author, the target text readers and the revisers' personal preferences.

The theoretical basis for these six categories can be traced back to Thaon and Horguelin's (1980) revision guide (in Shih 2006:297-298), which distinguishes five categories on which a reviser can base his or her checking of the translation, while the adoption of such a category as a norm during revision would also give an indication of priority or loyalty. The following five categories make up Thaon and Horguelin's guide:

1. Accuracy, which indicates semantic accuracy and includes detecting mistranslation, incoherence, omissions, use of imprecise language and failure to render the nuances contained in the source text. (The use of this category as a norm during revision would therefore imply loyalty to the source text.)

2. Correct usage, which includes syntax, grammar, barbarisms as well as typographical errors, for example spelling and punctuation. (The use of this category as a norm during revision would therefore imply loyalty to the target language.)

3. Transparency, which means that the translation must read like an original, being idiomatic, clear, concise, cohesive and free from repetitions. (The use of this category as a norm during revision would therefore imply loyalty to the target language or the target text.)

4. Tone, which means that the same levels of diction must be maintained, such as formality or even humour. (The use of this category as a norm during revision would therefore imply loyalty to the source text.)

5. Audience appropriateness, which indicates consideration of cultural content with regard to the target audience. (The use of this category as a norm during revision would therefore imply loyalty to the target audience or the reader.)

These categories mentioned by Thaon and Horguelin (in Shih 2006) are also reflected in the 12 revision parameters listed by Mossop (2010:125), which are grouped together in four categories:

1. Transfer issues, which would imply that the source text is prioritised.

2. Content issues, which would imply that the source text is prioritised.

3. Language issues, which would on a first level imply that the target language is prioritised but on a second level also the target readers - with reference to tailoring, meaning that the language should be "suited to the users of the translation and the use they will make of it".

4. Presentation issues, which would imply that both the target language and target text are prioritised.

Drawing on these two revision checklists, the categories target language, source text, target text and target text reader were established as possible priorities during revision. The category of author was added based on Hatim and Mason's (1990:16) distinction between author-centred, text-centred and reader-centred translation. Like Mossop, Hatim and Mason (1990:17) see translation as involving a conflict of interests, but instead of Mossop's suggestion to shift loyalties during the subsequent phases of translation and revision, they take the role of genre into account when explaining how this conflict might be handled in practice. Author-centred translation is often found in the case of literary translations when translators base their translation choices on the symbolic capital of the source text author. In the case of legal contracts, however, the source text author is of no importance but the source text itself needs to be replicated as closely as possible. In this case, a text-centred approach would apply. In the case of persuasive texts such as advertisements, neither source text author nor source text as such is important. During this translation process, the target text needs to communicate with the reader and therefore the translator would use a reader-centred approach (Hatim and Mason 1990:16-17).

Personal preference as a sixth category was added after an initial coding phase when it became apparent that certain choices were not made on the basis of loyalty to any of these factors but were based solely on personal preference (e.g. when an agent says, "I feel that this should be ...").

The coding was undertaken twice. After the first round, the categories were adjusted to include the sixth category that had emerged during the coding process. The second round of coding was undertaken by a different researcher, and the results were checked in a last round by the authors.

 

4. Results

The results are presented in Table 4 in a quantified form as the number of comments expressing different priorities. These results were used in an inferential analysis to reveal possible significant and nonsignificant differences among the six factors (see Figure 1).

 

 

5. Discussion and interpretation of numerical data

Inferential statistics show that the agents are loyal to the target text language, with only the mean for loyalty to the language being significantly different from the means for the other five factors (Figure 1, Table 5 and Table 6).

 

 

 

 

Overall, 46% of the comments revealed loyalty to the language. This preference is also reflected in the rather high scores for target language in Text 1 (57.4%) as well as in Text 3 (63%). The overall average of 46% is a consequence of the lower percentage for Text 2 (20.8%), where loyalty to the source text is dominant among the agents. As set out in the methodology section, loyalty to the target language would imply that an agent is concerned with matters such as correct usage of syntax, grammar, spelling and punctuation. Meanwhile, the translated product should read like an original text, meaning that the idiomaticity of the target language is an important norm. As the inferential statistics show, loyalty to the target language is the only factor that differs significantly from the other five factors, meaning that for the three processes studied, language issues received the most attention from the agents.

Loyalty to the source text itself appears in the second place with an overall percentage of 17.3%, which does not differ significantly from the other factors, excluding target language (refer to Figure 1). The percentages for Text 1 (13.6%) and Text 3 (11%) are lower than this average, while the score for Text 2 is higher (26.5%). Loyalty to the target text features in the third place, with an overall score of 12.2%, which is not significantly different from the other factors, excluding target language. For this factor, the highest score was again obtained for Text 2 (18%), while Text 1 came second (14.2%) and Text 3 third (6%). From this, it seems as if the scores for Text 2 for these two factors fall out of range, being higher than average, suggesting that loyalty to the source and target texts may have played a larger role than in the case of the other two texts.

Overall, the scores for loyalty to the author (8.4%), reader (8.1%) and personal preference (8%) are all low and inferential statistics again do not show significant differences. Only the results for Text 2 might be seen as slightly out of the range, with 12.4% for loyalty to the author and 16.3% for loyalty to the readers.

5.1. Text 1

Among the agents involved in the production process of Niggie/To hell with Cronjé, the editor (70%), the translator during her revision of the author's revised version (48.5%) and the author (33.3%) all give preference to language matters. The translator, in her initial translation notes, gives preference to the source text (45.4%), but language matters are not neglected, with 36.4% of her comments dealing with them. The reviser, however, seems to be specifically loyal to the target text (40%), while language matters in this case receive less attention (26.7%).

The editor's notes (J5) that were produced towards the end of the production process clearly indicate a strong focus on correct usage as well as idiom. Typical comments by the editor include the following:

J5.4 Off-saddle: I know this is a perfectly correct translation, and there is such a thing as off saddling, but I don't think it's generally used in English. I think it's more common to talk about dismounting. [Idiom]

J5.5 Like and As: I have always understood that like is used when comparing nouns (one thing is like another), and that when it is verbs, the doing of things, that are being compared, one uses as. So, on p. 8, and again on p. 21 Like every other evening should be As on every other evening. p. 30 sleep as never before, also pp. 100, 128,190, and 194. But like is creeping in more and more frequently, and seems to be becoming common usage. I notice it all the time, and it jars, but I am probably just a nit-picking old fuddy-duddy. [Language usage]

J5.17 Large of stature: stature means height. So, either tall of stature, or of large build. Pleasure(s) of the senses: there is surely more than one type of sensual pleasure? (Nit-picking again.) [Language usage and grammar]

J5.81 Between them stands a stranger - verb must be singular. [Grammar]

The editor also made notes that reflect loyalty to the stylistic feel of the target text, for example when she comments on the use of the English versus the Afrikaans word although she finds the English word perfectly acceptable:

J5.7 Minister/predikant: For what it's worth, I quite like predikant, which sort of adds to the flavour, but there is nothing wrong with minister. [Target text]

However, she also reveals her knowledge of the source text and its intertextual echoes when she suggests a change that would retain the rhythm and sound of the intertext that is invoked:

J5.15 Carry me back to the old Transvaal: Bring is the literal translation, and it scans better, and fits in with the tune for anyone who knows the song. [Source and target texts]

The translator's comments made during her revision of the author's revised version (J4), before the draft was sent to the editor, also show that language matters are emphasised during this phase. Idiom receives special attention, together with grammar, as the following typical comments in this section show:

J4.7 'kotching' - dink jy ons kan wegkom daarmee? Of moet mens maar liewer praat van 'throwing up'? Ek het baie getwyfel, maar sien jy het dit gelos. Dit kom wel in die Dictionary of SA English voor. Ek sluit dit in die glossary in - indien dit nie gebruik gaan word nie, moet ons dit net uithaal.

'kotching' - do you think we can get away with it? Or should one rather talk of 'throwing up'? I was much in doubt, but I see you left it. It is however given in the Dictionary of SA English. I include it in the glossary - if it is not going to be used, we just to take it out. [Idiom, although the reader is kept in mind when the expression is added to the glossary.]

J4.18 Ek dink nie mens kan 'accursed' so gebruik nie - gewoonlik as adjektief. Ek het dit sommer net uitgehaal.

I do not think one can use 'accursed' like this - usually as adjective. I simply removed it. [Grammar]

However, the notes made by the translator earlier in the process while still drafting the initial translation show that loyalty to the source text is predominant at this stage. For example, she queries certain facts in the source text that seem to be contradictory:

J1.8 Ben en Reitz beide op perde. Het hulle dan nie een perd en 'n muil gehad nie?

Ben and Reitz both on horseback. Didn't they have just one horse and a mule?

The author's notes that were sent to the translator and reviser (J2), after she had revised the draft translation, also indicate that language matters are closely considered (33.3%), although the percentages of comments indicating loyalty to the source (8.3%) and target (16.7%) texts as well as the author (25%) might mean that the author strives to keep these aspects in mind as well. Her comments include those in which she demonstrates loyalty to her own style, such as the following:

J2.5 In the Afrikaans text I use hy sê (he said) virtually with no variation. I have often changed words like he declares, remarks, muses, etc. back to he says, as this is characteristic of my style. But it is a difficult one. Please advise. [Author]

She points out issues of language usage as well, as in the following comment:

J2.6 Abbreviated forms such as can't (for example) should be used in direct and reported speech, and the unabbreviated form such as cannot in straight text. It is possible that there are still some inconsistencies in the text, although I have checked. [Language usage]

At other times, the author reveals loyalty to the source text in that she wants to retain the same cultural congruence in the target text as in the source text when she points out the inconsistency in the translation of the names of the characters:

J2.11 Names changed:

Swartpiet > Blackpiet, Rooi Herman > Red Herman, Gif Luttig > Blighter Luttig, Sagrys Skeel > Sagrys Sadie, Grapjan > Jan Jokes, Stofman > Dustbag, Kinderpiel > Smallprick, Jakobus Waterval > Wagenaer.

This is really a problem, because why change Skeel and not Stilgemoed, for instance. I will need time to think about this. [Source text]

As already mentioned, the target text's texture was the main concern for the reviser (40%), as is evident in the comments addressing the author's concerns as expressed in J2.5 and J2.11:

J3.5 At first I thought I would agree that you should keep "he said" throughout. I am in favour of the simple style. But actually I think the way the translator has done it works quite well. The variations are not intrusive - they do soften or round out the texture a bit, but perhaps this is a good thing. One bit I like particularly is on p37 of the translation beginning "Ezekiel was raised by hand," Gert Smal declares suddenly. That whole section - with the words declares, comments, confirms - flows very smoothly and gives a nice feeling of irony. [Target text]

J3.10 I don't think it's a problem that you've only translated some of the names. I think "To Hell with Cronje" should be regarded as an original - if it's not an exact translation of Niggie it shouldn't matter at all. If you translated all the names to English you would lose some of the specifically Afrikaans feel that I think is very necessary. These are boere out on the veldt; we don't want them to sound like khakis. I also wouldn't worry too much about the meaningful names being lost. It doesn't really matter if we can't translate Stilgemoed for ourselves - perhaps one can add a line emphasising his normal restfulness/stillness. [Target text]

The researchers were surprised to find that loyalty to the reader was expressed in only 5.9% of the comments made during the production process of To hell with Cronjé, especially for the reason that the South African landscape with its unique geology (on which Reitz is an expert), together with the fauna and flora (on which Ben is an expert) feature abundantly in the text. Yet these indigenous references are not really treated in a systematic way so as to ensure that all readers (e.g. South African English readers or even an international audience) understand all the terms and references.

The translator, in her initial translation notes, points out that she did not translate all the plant names - some she found untranslatable. However, she seems concerned about the target text readers when she declares that she is always careful not to foreignise too much:

J1.9 Plantname - ek het sommiges vertaal, ander (die onvertaalbares) net so gelos. Dis ook 'n opsie om al die name in Afrikaans te los. (Dis wat ek op die ou end met This Life gedoen het.) Julle moet maar vir my kyk of dit werk soos dit nou is. Ek is altyd bang om te erg te vervreem.

Plant names - I translated some of them, but left the untranslatable ones. It is also an option to keep all the names in Afrikaans. (That is what I eventually did with This Life.) Please have a look whether it works like this. I am always afraid of foreignising too much.

Early in the production process, most comments that do serve the needs of the readers are those by the author, suggesting the inclusion of a glossary. Later in the process, when revising the author's revised version, the translator also refers to the inclusion of elements in a glossary (see comment J4.7 in this regard as well).

The reviser asks for explicitation once, although she uses herself as the point of reference to motivate the request:

J3.11 P 26 of the translation - I stumbled a bit over 'hottentotsbedding'. This is too unfamiliar in English. Perhaps use a more descriptive phrase.

Towards the end of the process, the editor reveals some concern for the needs of the readers in her comment on words that describe various hat shapes, although her choice is also influenced by notions of correctness and acceptability:

J5.23 I had never heard of a claw-hammer coat, but according to the dictionary there actually is such a thing, so I'll go along with that. But semi-tophat sounds like a made up word, and terribly clumsy, and besides, the Groot Woordeboek defines pluiskuil hoed as chimney-pot hat, which I think is a more acceptable and familiar term.

The reader's needs are therefore not prioritised in any phase of the production process, nor by any of the agents specifically. Overall, the production process seems driven by loyalty towards language issues and, to a lesser extent, the source and target texts.

5.2. Text 2

The strong presence of the author among the text production agents for My name is Vaselinetjie is conspicuous. While the author of Text 1 presented only one set of comments (J2), the author of Text 2 generated three sets of comments: those made after having revised and reworked the draft translation (K2), her answers to the editor's questions (K4) and her notes to the editor (K5).

Loyalty to the source text dominated this production process, as is demonstrated by the majority of comments (26.5%). The agents who show the most loyalty to the source text are the editor (41.4% in her comments after her revision/editing) and the translator (34.7% in her comments during the revision phase). This loyalty often conflicts with the author's vision of the target text - she has, in fact, produced an almost new version of the source text during her revision of the draft translation. Loyalty to the target language took a second place (20.8%), while loyalty to the reader (16.3%), the author (12.4%) and personal preference (6%) took subsequent places.

The editor, in her comments after she has worked through the author's changes to the draft translation, is rather explicit in her criticism of the reworked version. She points out that the author has in fact started to add sections to the target text so that the character of the source text is compromised:

K3.13 Op party plekke het Anoeschka aan die skryf gegaan en soveel bygeskryf dat daar nou 'n hele "boggel" in die storie is. Dit word skielik op daardie plek 'n heeltemal ander soort storie, wat op 'n ander manier werk as die res. Ek het in sulke gevalle bietjie teruggesnoei.

In certain places Anoeschka went on a writing spree and has in fact added so much that the story is now distorted. In such places it becomes a whole different type of story, which works in a different way. In such cases, I did a bit of trimming.

The editor's loyalty to the source text is evident when she explains why she could not bring herself to accept all of the author's changes:

K3.7 Wat Anoeschka se veranderinge betref: Baie daarvan is goed en 'n besliste verbetering, maar soms is dit vir my asof Anoeschka in 'n heel ander "modus" is as wat sy was toe sy die oorspronklike Vaselinetjie geskryf het. Ek kon haar veranderinge dus nie slaafs in die teks aanbring nie omdat party daarvan die boek skade sou aangedoen het. En omdat ek ná al die jare nog steeds MAL is oor Vaselinetjie, sou ek nie graag wil sien dat daar enigiets aan die boek gedoen moet word wat dit bederf nie.

Regarding Anoeschka's changes: Many are good and a definite improvement, but at times it is as if Anoeschka is in a totally different 'mode' to when she was writing the original Vaselinetjie. Therefore, I could not effect her changes without questioning them because some of them would have harmed the book. And because, after all these years, I'm still BESOTTED with Vaselinetjie and I wouldn't want to see anything done to the book that might ruin it.

The author responds to this comment with a lengthy explanation of her reworking of the source text, claiming that it is based on her intimate knowledge of teenagers in 2008, as opposed to their peers in the late 1990s:

K4.7 1. Kinders van vandag is baie meer BLOOTGESTEL: aan allerlei euwels/seks/geweld én tegnologie. 2. Kinders het vir my "vervlak." Hulle is blootgestel aan al hierdie verskriklike dinge in die wêreld, maar ek ervaar dat hul emosionele ontwikkeling amper nou "stadiger" plaasvind .... Dus, ek het die INTENSITEIT/kruheid? van sommige tonele "'n oktaaf hoër" gevat, maar terselfdetyd amper 'n paar "hoërskool" lawwighede en oordrewe beskrywings ingebring.

1. The children/youngsters of today are much more EXPOSED: to all kinds of evils/sex/violence and technology. 2. Youngsters, I feel, are much more shallow. They are exposed to all these terrible things in the world, but in my experience their emotional development has been slowed down ... Thus, I took the INTENSITY/crudeness? of certain scenes to the next level, but simultaneously I inserted a bit of high-school silliness and a few inflated descriptions.

This type of justification for the reworking of the source text is seen quite often in the author's comments, revealing loyalty to herself as the owner of the source text and her vision of the target text. At times the author is adamant that her changes are to be respected, based on her expert knowledge of young people in 2008:

K5.5 Die verwysings na beroemde mense wat die kinders of admireer of wil wees, is BAIE spesifiek so gekies. LOS DIT ASB. NET SO, al maak dit moontlik nie vir julle sin nie. Trust me, ek het deeglik my huiswerk gedoen.

The references to famous people that the youngsters admire or want to be have been VERY specifically chosen. PLEASE LEAVE THEM AS THEY ARE, even if they do not make sense to you. Trust me, I did my homework thoroughly.

The changes effected by the author are not limited to content but involve language use as well. The translator, in her notes produced towards the end of the process, expresses her disdain at the high frequency of Americanisms that appear in the author's latest version and feels that the South African flavour of the book is thereby compromised:

K6.3 Ek stem nie saam met die Amerikanismes waarmee die teks nou besaai is nie. I don't think it rings true in the South African context - al is die jonges ook hoe onder die invloed van televisie, ens. Sien bv. outta (p. 123); cussing (p. 57) - ek het nog nooit gehoor dat iemand die woord gebruik behalwe in 'n cowboy-fliek/boek nie. Ek hou ook absoluut niks van bootie nie en sal regtig verkies dat dit nie gebruik word nie. Kan ons dit met iets anders vervang, asb? Dissed (p. 100) - ek sal nog daarmee saamgaan, maar sou verkies dat dit nie so baie gebruik word nie. (Sien ook pp. 79; 115; 194.) En, ai, ek hou ook niks van homie nie (pp. 115, 131.) Dis so American gangster!

I don 't agree with the many Americanisms that now appear throughout the text. I don't think it rings true in the South African context - even though the youngsters might be exposed to television, etc. E.g. see outta (p. 123); cussing (p. 57) -1 have never heard the word used except in a cowboy movie/book. I also detest bootie and would really prefer that it is not used. Could we please replace it with something else? Dissed (p. 100) -1 could agree to that, but would prefer that it is not used so often. (Also see pp. 79; 115; 194.) And oh, I really dislike homie (pp. 115, 131.) It's so American gangster!

Loyalty to the target text readers can be seen only in 16.35% of the comments. The editor, however, often identifies with the reader and anticipates problems regarding meaning:

K3.2 Ek weet nie of jy weet hoe mens jouself op 'n manier "dom hou" wanneer jy 'n teks redigeer nie? Anyway, ek het myself "dom gehou" en toe kom ek agter die feit dat daar in die vertaling nou van "house mothers" en dan van "matrons" gepraat word, maak die deur wawyd oop vir misverstand. Daar word aan die begin van die storie 'n paar keer van die "house mothers" gepraat, en dan word daar skielik van die "matron" gepraat. En daar "verstaan" ek dit toe as sou daar 'n klomp "house mothers" wees met net een "matron" aan die hoof van al die "house mothers" . . . Ek het toe die "house mother/s" deurgaans in "matron/s" verander. Ek dink regtig dis duideliker so.

I don 't know whether you know how one plays "dumb" when editing a text? Anyway, I played "dumb" and then I realised the fact that the translation refers to the "house mothers", but then to the "matron", leaving room for confusion. At the beginning of the story there are references to "house mothers" and then suddenly the "matron" is introduced. So I "understood" that there are a number of "house mothers" with only one "matron" at the top ... So I changed all the "house mother/s" to "matron/s". I really think it is clearer like this.

In another instance, the editor draws attention to the fact that the spatial description of the setting (the hostel) does not add up (and might be a source of confusion for the readers). She asks the author to provide a detailed drawing, which can be cross-checked by the proofreader.

5.3. Text 3

The comments that were generated during the production of In bushveld and desert: A game ranger's life show that the process as a whole was dominated by loyalty to the target language (63%). Loyalty to the source text takes the second position with 11%, although this loyalty seems to taper off as the production process evolves. (It does, however, become rather important again in the very last phase of revision by the author, when a score of 33.3% is seen.)

The comments regarding the structure of the compilation show loyalty to the target text, but this is to be expected during this phase of the production process. It is also during this phase

that some loyalty to the readers is seen (33.3% of the comments in L5.1, L5.2 and L5.3), for example when the structure of the target text is negotiated with the reader in mind:

L5.2 Ivm bg 2 stories: ek sal baie graag wil hê dat Moment in die bundel bly (en ek veronderstel Christiaan ook), en ek wil die vlg voorstel: Dat Delusion VOOR Moment geplaas word, dan kry die leser die hele agtergrond ivm die weermag/game ranger subterfuge.

Regarding the abovementioned 2 stories: I would like to keep Moment in the collection (and I suppose Christiaan as well), and I would like to propose: That Delusion is placed BEFORE Moment, then the reader has the background regarding the army/game ranger subterfuge.

The comments relating to the revision process, however, are indicative of a rather consistent concern with language matters. It is also noteworthy that the category of personal taste shows higher scores than is the case for texts 1 and 2: 25% of the comments made by the translator after having revised the compiler's edit and 26.3% of the comments made by the translator after having revised the editor's revision indicate choices that are motivated by personal taste.

The proofreader's comments at the end of the revision process show the highest level of loyalty to the target language (85%) - in line with the strong focus on the gatekeeper function that is to be expected during the proofreading stage. Comments such as the following are typical during the second last phase in the production process (with the author's final notes as the last phase):

M6.5 p. 76: 'larder' is a very English word - I don't think I have ever heard the word referring to a South African one - we would call it a pantry.

However, the earlier stages of revision are also dominated by a concern with target language: 76.8% for the author's comments on the translator's translation notes, 63.9% for the translator's comments on the compiler's revision, 59.3% for the translator's initial notes, 56.6% for the translator's comments on the editor's revision, 52% for the author's last comments and 50% for the translator's notes on the additional translation.

The translator's initial notes set the tone of her consistent loyalty to the target language that is to follow:

M1.50 Detour: Ek twyfel nou daaroor of hierdie storie hoegenaamd in die teenwoordige tyd vertel moet word. Werk dit te midde van al die ander verlede tyd vertellings? Ek het op die ou end begin 'faff en verander aan tenses. Nou het ek 'n vars oog nodig. Mens sou dit ook ewe maklik in die verlede tyd kon vertel.

Detour: I doubt now whether this story should be told in the present tense at all. Does it work among all the other past-tense narratives? In the end I started faffing and changing tenses. I need a fresh eye now. One couldjust as easily tell it in the past tense

The translator's comments on the compiler's revision are dominated by loyalty to the target language, and the following comment is typical of her revision during this phase:

M4.4 Hailstorm: p. 46: Tweede par. - Suzette het die tense verander. "On the opposite side, north of Nsikazi, are (ivp were) Khandizwe and Matjulu ..." Ek stem nie saam nie. Hoewel die par. in die present tense begin, verander die tense die oomblik dat jy sê "Our camp lay in its shade." Daarna beskryf jy die omgewing rondom die kamp soos dit was, nie soos dit is nie. Daarom past tense.

Hailstorm: p. 46: Second paragraph. - Suzette changed the tense. "On the opposite side, north of Nsikazi, are (instead of were) Khandizwe and Matjulu ..." I don't agree. Although the paragraph starts in the present tense, the tense changes the moment you say "Our camp lay in its shade." After that you describe the camp surroundings as they were, not as they are. Therefore, past tense.

Similarly, the translator expresses concern with language matters in her comments on the editor's work:

M5.5 p. 47: Skielik twyfel ek: Ek het gesê 'Disappointment lay heavy inside me.' Kan mens so sê? Moet dit nie tog maar heavily wees nie?

Suddenly I'm in doubt: I said 'Disappointment lay heavy inside me.' Can one say that? Shouldn't it be heavily, after all?

Next to loyalty to the target language, the translator's notes in this phase also reflect a strong inclination to express personal opinion (25%) and the words "I do not like" are rather common:

M4.3 Routine patrol: p. 32: Die laaste sin - ek het doelbewus "Then he laughed" gekies omdat dit 'n sterker einde is. Ek hou nie van "started laughing" of "started to laugh" nie.

Routine patrol: p. 32: The last sentence - I deliberately chose "Then he laughed" because it is a stronger ending. I do not like "started laughing" or "started to laugh".

This expression is also found in the subsequent phase when the translator revises the editor's work and comments:

M5.14 p. 64: Ek hou nie van 'Boet was open-mouthed' nie. Miskien eerder: 'Boet stared, open mouthed'?

I do not like 'Boet was open-mouthed'. Maybe rather: 'Boet stared, open mouthed'?

Loyalty to the author and his specific style can also be deduced from the translator's comments after having worked through the compiler's revision (11.1%), although it is less than the concern with the target language (63.9%) and personal taste (25%):

M4.9 Culling: p. 71: Ek het gesê: "There they would grow to adulthood and become "problem animals", trampling tourists who left the safety of their motorcars." Suzette se weergawe: "There they would . become "problem animals" - hating the scent of people, they will trample tourists who left the safety ..." Will moet asb. would wees. Ek voel Suzette verhelder te veel - sy stel iets duidelik wat die skrywer oorspronklik nie duidelik gestel het nie, want dit spreek vanself, en in die proses verswak sy die teks.

Culling: p. 71: I said: "There they would grow to adulthood and become "problem animals ", trampling tourists who left the safety of their motorcars." Suzette's version: "There they would ... become "problem animals " - hating the scent of people, they will trample tourists who left the safety ..." Will should please be would. I feel Suzette explicates too much - she clarifies things that were not stated as clearly by the author as they are self-evident, and in the process she weakens the text.

This same loyalty to the author's style, and to a certain extent the source text, becomes evident in the translator's summary of what is, according to her, the main problem with the compiler's revisions:

M4.24 Ter opsomming: Ek verstaan dat Suzette se veranderings hoofsaaklik daarop gemik is om die teks en die opeenvolging van gebeure duideliker te maak. Sy verander egter in die proses aan my vertaalstrategieë, bv. deur heelwat aan die skrywer se kenmerkende styl te verander en deur inligting by te voeg ter verduideliking/verheldering. Vir my is dit soms 'n geval van "stating the obvious". En is Christiaan se styl nie juis "understated" nie? By geleentheid voeg sy ook "he said" of "I said" by wanneer daar dialoog in die teks is, waar Christiaan dit in Afrikaans nie gedoen het nie. En sy skrap sinne/dele wat sy as nie tersaaklik beskou.

To conclude: I understand that Suzette's changes are predominantly aimed at clarifying the text and the sequence of events. However, in the process she modifies my translation strategies by, for example, changing a lot of the author's signature style, and by adding information in order to explain/clarify. For me this is sometimes a case of stating the obvious. And isn't Christiaan's style particularly understated? At times she also adds "he said" or "I said" to dialogue in the text, where Christiaan did not do so in the Afrikaans. And she deletes sentences/parts which she considers irrelevant.

Loyalty to the source text (19.8%) is seen in the translator's revision of her own translation when she asks:

M1.53 Boom/plantname: Hoe belangrik is konsekwentheid? Ek het (ter wille daarvan) feitlik deurgaans die name vertaal. Ek het maar meestal ge'google' vir die Engelse name.

Tree/plant names: How important is consistency? (For its sake), I translated the names almost throughout. I mostly googled the English names.

At this point she reveals that she is aware of the readers, although she refrains from making a final decision in this regard and instead asks the other agents for advice:

M1.53 Baie wildreservate/gasteplase gee die Afrikaanse en Engelse name op hulle webwerwe. Dit het my laat dink die oorsese/Engelssprekende gaste wil wel die Engelse name weet. [...] Wat dink julle?

Many game reserves/guest farms give the Afrikaans and English names on their websites. This made me think that foreign/English-speaking guests do want to know the English names. [...] What do you think?

This concern for the reader, however, remains limited (3.7% for the translator's initial notes and 10% for her additional notes).

When the author revises the translator's draft and comments on her translation notes, loyalty to the target language seems to dominate his to-the-point answers. In most of the cases his answers are aimed at simply resolving a language issue:

M3.23 Ek het geen idee wat 'speklap' is nie. Kon dit ook nêrens kry nie. Ek het maar volstaan met 'a roll of cloth'. Hulp sal waardeer word. Iemand het intussen vir my gevra of dit nie dalk 'shammy' (chamois) is nie, maw seemsleer. Mutton cloth

I have no idea what 'speklap' is. Couldn't find it anywhere, either. So I stuck with 'a roll of cloth'. Help would be appreciated. In the meantime, somebody has asked me whether it's not perhaps 'shammy' (chamois), i.e. chamois leather. Mutton cloth.

During the phase of final revision by the author, after proofreading has taken place, his short answers (but now with a certain insistence) reveal loyalty to the target language:

M7.5 bl 79 Lessermasked weavers NIE yellow weaver Lessermasked weavers NOT yellow weaver

M7.6 bl 90 Thermals NIE air currents Thermals NOT air currents

However, as stated in the first paragraph of this section, the author also refers back to the source text during this last phase, especially to point out any errors that might have been introduced during the previous phases:

M7.13 bl 144 Once they had cornered him NIE he had cornered them

Once they had cornered him NOT he had cornered them

 

6. Conclusions

The single most important conclusion to be drawn from the study of the documented practice of these three Afrikaans literary translation processes is that the processes of revision do not seem to be coordinated. No clear preference for the needs of the target text readers (as suggested by Mossop 2010) was displayed in any one phase of any of the revision processes. Gatekeeping activities were seen to dominate in the revision of all three texts, with language therapy playing a less marked and definitely not a delineated role. In fact, drawing from the results of our study, South African literary revisers do not seem to evince loyalty to the target readers; instead, they seem focused on the gatekeeping function throughout the revision phase.

For the Winterbach text, being firmly canonised in the strata of 'high' literature, the expectation might be that the author as well as the source text should be most loyally adhered to. However, the source text itself scored only 13.6% and loyalty to the author could be deduced from only five comments in total (2.9%), with three of these having been made by the author herself. In fact, the target text took the second place (14.2%), scoring slightly higher than the source text. The project remained dominated by a concern with language issues, constituting 58% of the comments. This echoes the overall findings.

In the case of Von Meck's prize-winning youth novel, one might have expected that the target readers would be the main source of loyalty and that the target text would be revised with a view to specifically serving their needs. Loyalty to the readers is expressed in 16.3% of the comments in My name is Vaselinetjie (the highest among the three texts), but ultimately the researchers observed that the author was loyal mainly to her own rewriting or adaptation of the source text, with the other agents finding themselves almost unanimously defending the source text.

Although Von Meck makes many references to her intimate knowledge of youngsters, she does not seem to be specifically concerned with the target text readers. Rather, she seems concerned only to create characters that might better fit the recreated context, but the real readers with their sets of expectations or defences seldom come into play as far as the author is concerned. In one instance, however, the translator takes into consideration the teacher who will be responsible for reading aloud some of the racier scenes, which might cause embarrassment in a classroom situation. In other instances, agents other than the author seem to have the reader's interest at heart, for example when the editor asks for a map of the hostel so that she and the proofreader can check the locality of different places in the story to help make sense of the setting for the readers. In the final analysis, however, the source text has the loyalty of the agents working on the process, acting as gatekeepers who seem to need to defend the source text at all costs against the changes imposed by the author during the revision process.

For the In bushveld and desert stories, a high degree of loyalty to the readers is to be expected - many of them are international tourists (clients of the author, often from the United States of America) who might be interested in learning the names of the many indigenous species found in Southern Africa - but with reference to the three processes studied, In bushveld obtained the lowest score regarding loyalty to the readers. Gatekeeping regarding language matters took the first place, with 63% of the comments dealing with the correctness of the language of the target text. Yet the possibility remains that this concern with the 'correct translation' is a way of serving the needs of the English-speaking readers, even though this is not explicitly stated. This applies to the other cases as well.

A number of topics for future research in this field have come to the fore during the course of this project:

1. It seems necessary to work on the terminology related to revision and editing since these two terms are often used interchangeably or in a haphazard way. It might even be necessary to ask whether a distinction between revision and editing is necessary at all.

2. The terms self- and other-revision need to be examined closely. Is it indeed the case that they are two discretely demarcated processes following each other in a linear way?

3. It is also evident that a set of relationships among the agents working on a translation is constructed - each agent apparently with his or her own defined role and power. The construction and realisation of these different roles could be studied sociologically, based on documented practice.

A question that can only be answered if this study were to be replicated is whether Mossop's earlier binary task division between translators and revisers rings true for technical translations. Such a task division might also be true for literary translation projects in other countries or in other publishing houses in South Africa.

 

References

Bakkes, C. 2004. Stoffel by die afdraaipad. Cape Town, Pretoria and Johannesburg: Human & Rousseau.         [ Links ]

Bakkes, C. 2006 (2000). Stoffel in die wildernis. Cape Town and Pretoria: Human & Rousseau.         [ Links ]

Bakkes, C. 2007. Stoffel se veldnotas. Cape Town and Pretoria: Human & Rousseau.         [ Links ]

Bakkes, C. 2008. In bushveld and desert: A game ranger's life. Cape Town: Human & Rousseau.         [ Links ]

Hatim, B. and I. Mason. 1990. Discourse and the translator. London and New York: Longman.         [ Links ]

Mossop, B. 2010. Revising and editing for translators. 2nd ed. Manchester: St. Jerome.         [ Links ]

Mossop, B. 2014. Revising and editing for translators. 3rd ed. London and New York: Routledge.         [ Links ]

Munday, J. 2012. Evaluation in translation: Critical points of translator evaluation-making. London and New York: Routledge.         [ Links ]

Nord, C. 1997. A functional typology of translations. In A. Trosborg (ed.) Text typology and translation. Amsterdam: John Benjamins. pp. 43-66.         [ Links ]

Scocchera, G. 2016. The sociology of revision: Results of a qualitative survey on the professional practice of translation revision for publishing purposes and its agents, and how they relate to what we know (or we think we know) about revision. Paper read at the Eighth European Society for Translation Studies (EST) Congress, 15-17 September, Áarhus, Denmark.

Shih, C. 2006. Revision from translators' point of view: An interview study. Target 18(2): 295-312.         [ Links ]

Spies, C-M. 2013. Die wisselwerking tussen die agente betrokke by die publikasieproses van literêre vertalings. Unpublished doctoral dissertation, Stellenbosch University, Stellenbosch, South Africa.         [ Links ]

Von Meck, A. 2004. Vaselinetjie. Cape Town: Tafelberg.         [ Links ]

Winterbach, I. 2002. Niggie. Cape Town: Human & Rousseau.         [ Links ]

Winterbach, I. 2007. To hell with Cronjé. Cape Town: Human & Rousseau.         [ Links ]

Winterbach, I. 2010. To hell with Cronjé. New York: Open Letter.         [ Links ]

 

 

1 "Revising is that function of professional translators in which they find features of the draft translation that fall short of what is acceptable, as determined by some concept of quality [...], and make any needed corrections and improvements" (Mossop 2010:109).
2 "Editing means reading a text which is not a translation in order to spot problematic passages, and making any needed corrections or improvements" (Mossop 2010:17).
3 We do not refer to 'loyalty' as Nord (1997:48) uses it in functionalist theory (as "an interpersonal category referring to a social relationship between individuals"). We rather use the word in the generic sense of 'commitment' and 'allegiance'.
4 Mossop, however, never proposes this as the only solution, although he seems to be favouring this strategy. It should also be mentioned that Mossop (2014) has omitted this proposed binarity from the third edition of Revising and editing for translators.
5 The authors would like to express their gratitude to Dr Carla-Marié Spies-Gaum for the use of the set of data collected by her and published as appendixes to her PhD dissertation "Die wisselwerking tussen die agente betrokke by die publikasieproses van literêre vertalings" (Spies 2013).
6 The translator's initial translation notes are seen as indicative of a self-revision process and are therefore included.
7 Some comments were coded two or even three times, and in such cases they were all added to obtain 'n' for the specific agent.