Continuing the conversation among the words: Breyten Breytenbach’s translation strategy in Oorblyfsel/ Voice Over

Breyten Breytenbach’s bilingual collection of poems, Oorblyfsel/ Voice Over (2009), consisting of twelve poems published with the Afrikaans version on the left page and the corresponding English version on the right, shows some notable differences between the two versions of each poem. This article examines the unusual translation strategy of “re-tracing” (Odendaal 2011) that Breytenbach used to produce this collection of poems. Three concepts from literary theory, the ‘contact zone’ (Pratt 1987, 1992), the ‘in-between’ (Bhabha 1993) and the ‘remainder’ (Lecercle 1990), are discussed as tools which can be used to understand this strategy. In light of these three concepts, the two versions of the twelfth poem, simply named 12, are analysed in order to illustrate and understand the effect of Breytenbach’s strategy on the interpretation process that this collection of poems invites the bilingual reader to take part in.


Introduction and context
Oorblyfsel/ Voice Over (Breytenbach 2009), a bilingual collection of twelve poems published with the Afrikaans version on the left page and the English version on the right, was written as a eulogy to Breyten Breytenbach's deceased friend, the famous Palestinian poet, Mahmoud Darwish. In the Note at the back of the publication -also appearing bilingually like the poems - Breytenbach (2009:61, 63) says that he began writing the poems shortly after Darwish's death as a "continuing dialogue" with the poet and his work.
Translation is an important component of this dialogue. Breytenbach (2009:61, 63) makes it clear that he does not know Arabic and that he has only had access to Darwish's poetry through the content of English and French "approximations" and the rhythm and sound of Darwish's Arabic during readings. He therefore calls the collection a "'collage' [of] transformed 'variations' of [Darwish's] work." In the same way that he does not use the word "translations" to refer to his reworkings of Darwish's work, he does not use the word "translation" to refer to the relationship between his own Afrikaans and English "versions" or "efforts". This hesitance does not mean, however, that translation does not stand central to both of these relationships. In fact, Breytenbach (2009:61, 63) describes translation as a journey of proceeding from one language to the other, and equates it to his journey of "moving between" Darwish's verses. His hesitance rather points to the unusual nature of the relationship between Darwish and Breytenbach, their work and the languages and utterances at play here.
The conversation between Darwish and Breytenbach is, in fact, echoed in the relationship between Breytenbach's own Afrikaans and English versions. According to Breytenbach (2009:63), he wrote the Afrikaans versions first with the English versions "grow [ing] retroactively" from his Afrikaans "efforts". The versions can "not be properly described as 'translations'" because "[a]t times an echo or an association presented by the English possibilities opened a new way back into Afrikaans". Just as Breytenbach describes his poems as a "continuing conversation" with (the deceased) Darwish, he sets up the relationship between the Afrikaans and English versions as one of continuing interchange and interaction. Even though the Afrikaans and English versions are in most cases lexically equivalent to each other, Breytenbach's strategy does result in several cases where there are conspicuous differences between the versions in the two languages. These differences include aspects such as the number of lines, spacing and differences in the meaning of some corresponding words. Among these differences, one in particular forms the focus of this article: there are numerous examples of word pairs that differ semantically but are similar phonetically (how they sound) and graphically (the way they appear on paper).
Breytenbach's choice to translate certain words phonetically and/or graphically rather than semantically is an unusual strategy. Words from a different language that look and/or sound similar to the original but have different meanings are usually called "false friends" (Roca 2010) because they are often wrongly used as lexical equivalents by translators or second language speakers. It is not surprising that in the first review of Oorblyfsel/ Voice Over, Hans Pienaar (2009), in an otherwise positive review, finds the English to be surprisingly bad and even suggests that the Afrikaans reader would be better off ignoring the English. However, and as Pienaar himself points out, the English in the title of the collection should point the reader to the fact that both languages are crucial to an interpretation of the poems. The fact that Breytenbach repeatedly uses this strategy throughout the collection of poems also points to the fact that this choice of strategy was deliberate.
In an article focusing on this choice by Breytenbach, Odendaal (2011:300-306) names the strategy of translating sound and graphics rather than meaning "re-tracing" (oortekening). Grounding Breytenbach's unusual translation strategy in Derrida's work on translation, the nature of self-translation and Breytenbach's love of wordplay, Odendaal concludes that through re-tracing Breytenbach intimately intertwines the Afrikaans and English versions despite the obvious differences in meaning. In a review written about a year earlier, Oppelt (2010) reaches a similar conclusion: [T]he glaring differences between the Afrikaans and the English bodies of the same pieces are almost put side by side for the reader and listener to pick up on, and the space between is exactly where the reader and listener is invited to experience some very intimate thoughts and reflections.
In what seems like a response to Pienaar's assertion that Breytenbach's English is "bad", Oppelt further comments: "Yes, the English pieces are bad, but bad has seldom worked so well." Oppelt's comments highlight two obvious but important points. It is the bilingual layout that enables the reader to notice the differences between the versions. In fact, it is made very difficult for the reader to simply ignore the one language or the other. However, these differences in meaning are only noticeable for a bilingual reader -it would be lost on someone who can understand only Afrikaans or only English. Rose (1997:2) uses the term "stereoscopic reading" to refer to the simultaneous reading of two translations of one source text or the simultaneous reading of the source and the target texts. The bilingual layout as well as the translation strategy of re-tracing enables -in fact, invites -the bilingual reader to read the two versions stereoscopically, going from one version to the other and considering the two together in the process of interpreting what is being offered.

Three concepts from literary theory
The key aspects determining the nature of translation in Oorblyfsel/ Voice Over, the bilingual layout and the strategy of re-tracing which results in stereoscopic reading, are aspects which are not found in most other translation products. The majority of translation products are published and read separately from the source text. Because of this, most translation theory is focused on the distinction between source text/ language/ culture versus target text/ language/ culture, a distinction which becomes less helpful in this case as it is exactly the blurring of the lines between source and target that defines the nature of this translation and the resulting poems.
Rather than staying bound to translation theory (Pym 2010:166), this study operationalises concepts from literary theory as "thought apparatus" (denkgereedskap -Anker 2007:3) to examine the unusual translation relationship in Oorblyfsel/ Voice Over and the stereoscopic reading process in which it results. The three literary concepts that are used are the 'contact zone' (Pratt 1987(Pratt , 1992 the 'in-between' (Bhabha 1993) and the 'remainder' (Lecercle 1990). These three concepts describe different aspects of the translation relationship and the stereoscopic reading process, and they therefore complement each other. Even though previous applications of these concepts to translation products and the translation process have been criticised, and often for valid reasons, this article considers these critiques in the hope of showing ways in which these concepts can be valuable if they are applied in a specific way, as in the case of this specific translation phenomenon.

2.1
The contact zone Pratt (1987Pratt ( , 1992 uses the term "contact zones" to refer to "social spaces where disparate cultures meet, clash and grapple with each other" (Pratt 1992:4) and specifically to provide a way in which to approach the "fractured reality of linguistic experience in modern stratified societies" (Pratt 1987:51). Because relationships in the contact zone are complex and because no person or group has an assigned place in a hierarchic understanding of language and language users, the language use in contact zones is different from the way it is used within the same culture or grouping. These differences can only be understood if the contact is kept in mind.
Pratt does not focus specifically on translation but rather on the manifestation of power relationships in language use in general. However, because of this, she often compares different languages or dialects and their users. Her theory therefore seems to lend itself to translation studies. In fact, it seems very easy to argue that translation automatically creates a contact zone. The well-known translation theorist Lawrence Venuti (2000:477), for example, refers to translation as "a linguistic 'zone of contact' between the foreign and translating cultures, but also within the latter" (Venuti 2000:477). However, unlike Pratt's original use of the term, Venuti's application of her term does not make clear where the contact zone is situated (whether in the translation process and/or the product and/or the persons), which persons form part of the contact zone and how the contact zone specifically influences language use.
As Pym (1996Pym ( , 2001 convincingly argues, because translation products are clearly positioned within a specific culture and language, they are often less hybrid than original texts and therefore they do not necessarily create or function in a contact zone. Pym shows that in most translations most traces of the source text, and especially the source language, are removed in order to produce target texts that seamlessly fit in with the target language and the target culture. This type of translation affirms the homogeneity of the target language and keeps languages and cultures apart and different rather than bringing them into contact.
However, in Oorblyfsel/ Voice Over, unlike most other translations, there is a very clear and visible contact zone between the two languages in the text itself because of the bilingual layout. The contact zone can therefore be applied very specifically to the contact between the Afrikaans and English in this collection of poems. Because the two different versions of the poems create and function in this contact zone, the language use and the translation relationship are different than what they would have been if the versions were published separately. The contact zone and translation influence each other in two ways: the contact zone exists because of translation, and the natures of the translation process and products are affected because they function within a contact zone.
Re-tracing is possible because the similarity in form links the versions to each other. However, at the same time, the difference in meaning between the re-tracings is highlighted by the contact zone. This difference causes conflict between the two versions. The conflict of the contact zone makes it necessary for the reader to "grapple with" (Pratt 1992:4) the contradictions produced by differences in meaning, resulting in a variety of interpretations and meanings. The concept of the 'contact zone' is useful in describing the nature of translation in this collection of poems because it provides a way in which to make sense of how the two languages function together, rather than taking one of the languages out of the relationship, as proposed by Pienaar (2009). Reading the two versions together makes it possible for different and varying meanings to be produced.

The in-between
The in-between is one of several denotations that Bhabha uses to gives utterance to cultural difference. He explains the concept behind the term in an article entitled Culture's In Between (Bhabha 1993) that discusses the development of different cultures in relation to each other and considers which conception of culture is necessary to understand this coexistence of cultures. The article describes the postcolonial and postmodern realities because of which not only people and cultures but also conceptual frameworks have to be shifted and re-established. The type of cultures and world views that develop because of these shifts "is indeed something like culture's 'in-between', baffingly alike and different" (Bhabha 1993:167). The term "inbetween" is used in an attempt to give utterance to the reality that a person or group of people can simultaneously belong to two or more cultures and therefore also not completely belong to any one of those cultures (Bhabha 1993:167).
It is important to note that the "in-between" as a word can refer to time as well as space, and both these meanings are active in Bhabha's application. Some of the other words and phrases that he uses in conjunction with the in-between refer to both time and space: "in the moment of transit" (Bhabha 1994:1), "revisionary time" (Bhabha 1994:7) and "time lag" (Bhabha 1994:240, Bhabha in Mitchell 1995 refer to time while "here and there, on all sides, […] hither and thither, back and forth" (Bhabha 1994:1) as well as "in-between spaces" (Bhabha 1994:1), "intervening space" (Bhabha 1994:7) and "supplementary space" (Bhahba 1994:163) refer to space. He also often uses the words "space and time" together (e.g. Bhabha 1993:167 and1994:1). Both the temporal and spatial aspects of the word should therefore be kept in mind when using the term.
The in-between is useful when examing Oorblyfsel/ Voice Over because it can be used to describe the time and space of interpretation when and where the conflicting meanings as well as the similarities between the versions are simultaneously present. Because of re-tracing, the differences between the versions cause different interpretations to be made when the versions are considered together than if they were read separately. The interpretation(s) which can be made are also not set in stone for the reader. The in-between is the interpreting time and space when and where the reader can consider and explore the tension between the versions with varying, differing interpretations.
Even though Bhabha's theory, including the in-between, has been and continues to be used to describe and analyse translation products and processes, strong criticism from some translation scholars has been brought against this appropriation. The reason for this is that Bhabha often uses the word "translation" in his theory, but he uses it metaphorically, applying it to culture rather than language and discussing experiences rather than texts. (See, for example, his interview in Mitchell (1995) as well as Bhabha (1994:6, 162-163, 241, 234).) One of the strongest critiques brought against the use of Bhabha's theory in translation studies has to do with the fact that his version of cultural translation is not necessarily applied to texts or languages (Pym 2010:161, Trivedi 2007. Moving away from the text and from two languages is firstly problematic because the question has to be asked whether cultural translation can then be classified as translation at all. The translator and translation theorist Harish Trivedi (2007:286) argues passionately for retaining the meaning of translation inherently grounded in the interlingual transfer of texts: "[W]hat Bhabha, with his usual felicity, has in another context called 'non-substantive translation', […] one could perhaps go a step further and, without any attempt at matching felicity, call it simply non-translation". Pym (2010:165) suggests that studies in cultural translation should rather be called "intercultural studies", a term which he views as an overarching field of which translation forms a part (Pym 1998:199-201).
Because cultural translation is not the same as (interlingual textual) translation, a further point of critique can be raised in that what cultural translation propounds is not necessarily valid for (interlingual textual) translation. Tymoczko (2003:182-183, 185, 195-196) and Baker (2005) argue that the "in-between" implies that the translator is in a neutral space without any cultural prejudices, an implication they are strongly opposed to because translators are people who cannot free themselves from their cultural position. However, this interpretation of the in-between as implying a neutral space 1 is not accurate according to Bhabha's use of the concept. The in-between does not only refer to space but also to time, the time when the dialectic thesis and anithesis co-exist in tension but before the synthesis has taken place. Accordingly, and secondly, the in-between is definitely not a neutral space. It is a space and time where different conflicting meanings co-exist, tying in with what Baker (2005:12) says: "[W]e need to recognize and acknowledge our own embeddedness in a variety of narratives". Bhabha's conceptualisation of the in-between is his attempt to explain the process when people or cultures are confronted with narratives that conflict with the narratives they already believe in, or when they find themselves in two or more conflicting narratives.
Tymoczko's and Baker's critiques point out that translators work within a specific ideology, geography and time. Translators can only move in the in-between when they are confronted with the conflicting aspects of the languages and cultures of the source text versus those of the target text. As soon as they make a decision about how they are going to resolve that tension and, even more so, when they execute that decision, they bring about a synthesis and move out of the in-between. To position all translators in the in-between is not only an inaccurate reflection of reality, as Tymoczko and Baker point out, it is also a misinterpretation of Bhabha's use of the term.
The in-between can therefore not necessarily be applied to all translations. However, in this article it is being applied to a very specific translation strategy in a very specific text. The inbetween offers a way to describe the process through which the bilingual reader reads and interprets the examples of re-tracing in the bilingual text. Because the concept is being applied to a text and to the relationship between two languages, Trivedi's (2007) concern about "nontextual" and "monolingual" translation is also refuted. The in-between, as well as other concepts from literary theory, can therefore be useful in describing and analysing the linguistic aspects of texts, and not only metaphorically. It should however be done critically and with good judgement.

The remainder
The remainder is the main focus of Lecercle's book, The Violence of Language (1990), which is a philosophic and linguistic examination of language utterances that do not fit into the rules and structures of conventional linguistics. He finds in these utterances the manisfestation of "another side to language" (1990:6) from that which is known linguistically, and the remainder is his attempt at making sense of the (seemingly) superfluous, incomprehensible, exceptional, ungrammatical, irrelevant, ambiguous and unclear aspects of language as found in real language utterances (1990:52). These aspects of language are not necessarily linked to lexical meaning. The remainder is manifested in the conflict between form and meaning, where the nonmeaningful parts of language find meaning.
The translation process foregrounds both the distinction between form and meaning as well as the fact that these two aspects influence each other. This distinction between form and meaning is what makes Lecercle's remainder possible: the remainder allows for the distinction between form and meaning to be utilised so as to release the ambiguity and the signifying multiplicity of language. In Oorblyfsel/ Voice Over, the translation strategy of re-tracing emphasises this distinction because form is translated rather than meaning. The resulting difference in meaning between the two versions makes it possible for the respective versions to result in different interpretations. Because the versions are presented together, however, the possible interpretations are multiplied even more. The interpretation(s) of one version can not only be applied to the other version, but the different versions together make interpretations possible which would not have been the case if only one version was considered by itself. The remainder refers to the ambiguous, superfluous and sometimes (seemingly) incomprehensible meanings that originate in the contact zone in-between the two versions in Oorblyfsel/ Voice Over.
Lecercle's remainder has already been applied to translation by Venuti (2000Venuti ( , 2002. Venuti uses the concept as the "(domestic) remainder" to refer to language utterances in a target text that contains dialects, discourses, registers and/or styles that belong to the linguistic, cultural, social and political context of the target language and that cannot be traced back to the source text. Venuti's view of the remainder as a manifestation of the domesticising effect of the target language (2002:219-20, 224) seems counter-intuitive to the basic meaning of the word "remainder", which suggests that it can be used to refer to culture-specific parts of the source texts which "remain" behind in the target text. According to this application, the remainder will be found in a target text when a translation strategy such as Venuti's "foreignization" (1995) is followed where certain cultural-and language-specific aspects of the source text are brought into the target text.
Both of these applications, with their differentiation between source and target, are not necessarily helpful when considering re-tracing as a translation strategy in the bilingual context of Oorblyfsel/ Voice Over. The way in which the concept is applied in this article therefore does not position the remainder within the source or the target language and text; rather, the remainder is positioned in the in-between of form and meaning in language and in the contact zone between the two languages.
Lecercle's description of the remainder in The Violence of Language offers a few additional concepts that can be useful in the description and analysis of Oorblyfsel/ Voice Over. Lecercle holds that even though the remainder can manifest in the most mundance utterances, it specifically flourishes in the language of "the delirious madman/patient" and "the inspired poet" (Lecercle 1990:6, 60, 108). The latter is more obvious: a poet "truly speaks language and uses it for the highest form of expression" (Lecercle 1990:108). Poets consciously use the ambiguity of language to result in unusual interpretations.
The language use of "madmen" can have the same effect, but without the same level of awareness, of course. Lecercle discusses two examples of such people. Brisset, a psychologically unstable French linguist, repeatedly analysed words etymologically in order to find (according to him) not only the history of words but of society, often with surprising results. For example, he "Brissetized" the word "Israelite" as y s'ra élite which means 'he will be the elite' (Lecercle 1990:62). Lecercle points out, however, that it is not only Brisset's overactive imagination which made Brissetizing possible; the ambiguity of language and of homophony in everyday life makes multiple analyses possible.
A second person that Lecercle (1990:63, 230) discusses is the Jewish American schizophrenic Louis Wolfson who could not stand hearing or reading words in his mother tongue, English. He therefore translated words immediately based on how they sound. Even though Wolfson took phonetic translations, also known as traducson (Lecercle 1990), to the extreme, many examples can also be found in everyday life. Lecercle discusses the French word contredanse which does not mean 'opposite dance' but is etymologically derived and phonetically translated from the English "country dance".
All of the above applies to Oorblyfsel/ Voice Over. Breytenbach can be viewed as the "inspired poet" who uses the ambiguity of language in his unusual translation strategy which results in unusual interpretations. The ambiguity of language, and specifically of how different languages sound (phonology), enables the reader of these poems to make multiple analyses of the differences between the versions.
Even though Lecercle does not attempt to situate the remainder within any specific theory, he does describe the concept several times using different theories. Two theories that he spends a lot of time on are Lacan's psychoanalytical theory of language and Deleuze and Guattari's materialistic theory of language (Lecercle 1990:33, 51). These two theories are, however, contradictory, as Lecercle (1990:42) points out. Psychoanalytic theories position language in the metaphysical terrain of abstract meaning. Deleuze and Guattari's materialism is opposed to psychoanalysis in that language for them is grounded in the material reality. This contradiction is not problematic for Lecercle. He holds that the remainder is situated exactly where the metaphysical and material aspects of language meet (Lecercle 1990:225) -in-between the metaphysical and the material aspects of language, so to speak. In the body of the poetry in Oorblyfsel/ Voice Over, this aspect of the remainder enables the tension between the metaphysical and material realities of language to be brought to the fore.
Another theorist to which Lecercle often refers is Heidegger, especially his phrase die Sprache spricht ('language speaks' -Lecercle 1990:5, 109-115). Lecercle frequently uses the phrase pair "I speak language" and "language speaks" to describe the remainder. In the case of Breytenbach's poetry, both these phrases are valid: there is not only a poet enunciator who speaks language and who has certain things that he wants to communicate; language is also allowed to "speak" in the sense that different readers can make differing interpretations of the same language utterances because the differences that are inherent in language are made visible through re-tracing. The remainder therefore not only functions in the thoughts of Breytenbach as he used language; it also functions for the readers who interpret the language utterances.
Because the remainder is an attempt to describe those aspects of language which cannot be described by rules and regulations, the remainder, by definition, also cannot be pinned down by rules and regulations. This fact notwithstanding, Lecercle (1990:121) proposes a tentative set of paradoxical rules for the remainder: "The remainder-work must and yet cannot be described in terms of rules". Even if they cannot adequately define the concept, these four rules are useful in illustrating the way that the remainder functions in Oorblyfsel/ Voice Over.
Lecercle's first rule is the rule of exploitation or flouting (Lecercle 1990:122-125), which states that the meaning found in language can be unleashed by exploiting the rules of language. In Oorblyfsel/ Voice Over, Breytenbach follows the convention of translating semantically for the most part but with the re-tracings, amongst others, he sometimes diverts from this convention. Because readers expect semantic equivalence, they compare the two versions with each other, which sets off the stereoscopic reading process. Breytenbach therefore utilises translation convention by exploiting this convention and the reading process it results in at times.
The second rule of the remainder is the rule of paradox (Lecercle 1990:125-128): a seeming contradiction can be present and/or true in language. In Oorblyfsel/ Voice Over, there are several examples where re-tracing results in seeming contradictions between the versions -where the different languages express opposites. However, these contradictions allow for different sides of the same experience or reality to be expressed.
Lecercle's third rule for the remainder is the rule of rhizome-work (1990:128-134). He derives the concept of the 'rhizome' from Deleuze and Guattari (1987). "Rhizome" is a term originally used in botany to refer to "a continuously growing horizontal underground stem with lateral shoots and adventitious roots at intervals" (Soanes andStevenson 2006:1234). Based on this image, Lecercle uses Deleuze and Guattari's five principles for how the rhizome works in language to describe the remainder.
The rhizome works according to the principle of connection: "a point in the rhizome can be connected with any other point. There are no fixed paths, as in a [Chomskyan] tree" (Lecercle 1990:132-133). Secondly, it works according to the principle of heterogeneity (Lecercle 1990:133). It does not enforce an ordering principle on language but only knows "a medley of contending dialects, heterogeneous styles, and idiosyncratic punning." Thirdly, it works according to the principle of multiplicity (Lecercle 1990:133). Because there are no set points or positions in the rhizome, there is no set place for speakers from where they can impose structure or order. Fourthly, the rhizome works according to the principle of non-signifying breaks (Lecercle 1990:133). A rhizome can break off or subdivide at any point and the different parts will simply start growing again. Finally, the fifth principle according to which the rhizome works is the principle of cartography (Lecercle 1990:133). A rhizome is not a tree which grows teleologically from a single root and trunk; rather, it is a map which "records the lie of the land" (Lecercle 1990:133).
In Oorblyfsel/ Voice Over, the connection between the Afrikaans and the English versions as well as the difference (heterogeneity) between them, which is brought about by re-tracing, allows a multiplicity of meanings between the versions. These meanings originate where links are formed in lexical meanings as a result of non-signifying breaks, allowing for various connections to be made. The landscape of these different meanings is not closely demarcated. According to the principle of cartography, the reader is invited to explore these differences while journeying between and through the versions.
In addition to the first three rules -exploitation, paradox and rhizome-work -Lecercle's fourth rule of the remainder is that of corruption and conservation. Language continually changes and therefore there is an inherent danger that language can destroy the values of individuals or cultures (Lecercle 1990:134). On the other hand, history is retained and kept alive through language. The paradox of corruption as well as conservation is ensured by the remainder. Even though the renewal that it brings to language is a threat, the same renewal also keeps language alive. Corruption and conservation therefore co-exist in this tension.
In Oorblyfsel/ Voice Over, the rule of corruption and conservation is not only found in how each version "threatens" the meaning of the other while also keeping it alive, but also in how the collection of poems reflects and comments on the relationship between Afrikaans and English outside of this publication. Afrikaans as a local, minority language is dependent on English, the world language, in terms of loan words and "access" to the world of technology, business, etc. to keep it alive (Ponelis 1993:120, Feinauer 2001. However, English also threatens the existence of Afrikaans. English as the world language is not completely independent of Afrikaans and other smaller, local languages. English is only kept alive by being able to adapt to and change in local environments, becoming localised with loan words and expressions. Even though local languages keep it alive, English is also threatened by becoming so varied across the globe that it runs the risk of losing its power as a single, uniform language. Oorblyfsel/ Voice Over reflects this complex relationship between Afrikaans and English, showing how the relationship is poised between being symbiotic and parasitic. It also reverses the power relationship of Afrikaans in relation to English by placing Afrikaans -the minor, local language -first, showing that Afrikaans can assert itself and that it does not have to be threatened to feel inferior by English's close proximity.

An interpretation of 12
The three concepts discussed in the previous section are useful to highlight and describe different aspects of the unique nature of Oorblyfsel/ Voice Over as a result of the translation strategy of re-tracing, the bilingual layout and the reading and interpretation process that it invites the bilingual reader to take part in. In order to illustrate how the theories discussed above play out in the reading and interpretation of the poems, the last poem in Oorblyfsel/ Voice Over, simply named 12 (pp. 58-59), is analysed. As mentioned in the discussion of the concepts, the nature of these poems, especially the two versions next to each other, makes possible a wide range of interpretations of even just one line of poetry. This analysis will therefore not attempt to unravel all the possible interpretations that can be made; it will only discuss some aspects of the poem, focusing especially on the examples of re-tracing and illustrating the stereoscopic reading and interpretation process.
12 is also difficult to interpret because, even though the rest of the poems do not always follow a clear narrative, this poem hardly has any narrative. It seems to paint a surreal montage of strange images that are woven together in and between the two versions. It also has the most examples of re-tracing per page of all the poems. The one page of the Afrikaans version has 15 words that are affected by re-tracing and its corresponding English page has 17. The average number of words affected by re-tracing throughout the rest of the collection is six per page. The high density of retracings in 12 makes it even more difficult to offer a single, linear interpretation.
In an attempt to portray the bilingual layout of the collection of poems, 12 is tabulated below, with Breytenbach's Afrikaans version on the left and his English version on the right. All words affected by re-tracing (in other words, corresponding words that look or sound the same in both languages but have different meanings) are underlined, with specific letters that look or sound the same in bold. The italics and ellipses in the poem are formatted as they appear in the original.
In order to accommodate readers who cannot read Afrikaans, I have translated the Afrikaans directly, word-for-word into English. My English translation, which is given in square brackets below each line of the Afrikaans version, makes no attempt at being poetic but only seeks to convey the meaning and as much of the linguistic structure of the Afrikaans as possible. My English translation is adapted for word order, however. If more than one meaning of a word was possible in the context of the Afrikaans, all possible meanings are given (separated by slashes). another day will break, a womanly day as song in gesture, green in greeting and squander. all will be unveiled beyond and besides what is past. water will slip from the stone's loins. no dust, no drought, neither defeat nor longing. and with noon a dove will nestle head under wing in the painted war wagon … if it cannot hollow a nest in the lovers' bed The montage of this poem paints a picture that is simultaneously utopian and ominous. In the Afrikaans version, images of beauty are interwoven with images of violence through re-tracing. The first line starts hopefully in the Afrikaans: 'n ander dag sal aanbreek ('another day will dawn'). Different from the male dispensation of war and violence, which the rest of the poems in the volume had described, this future day is a female one where the beloved extends her lovely time of beauty. The English version begins more ominously: "another day will inbreed, froglike and female". Like frogs in a pond, the other day will "inbreed" with all the previous days to become more and more the same. In this more ominous version, the Afrikaans beminde ('beloved') becomes the "bemoaned one" who has to "veil [her] lifelike beauty". Because sorrow is being expressed about her and because her beauty is "lifelike", it suggests that, in the English version, the beloved has already passed away. Her beauty will be veiled as she is being buried. The two versions express the tension between the lovely time of beauty in life versus impending death wanting to veil this beauty.
The second stanza of the poem starts in much the same way as the first. The first line of both the Afrikaans and the English versions, however, differ slightly from that in the first stanza. In the Afrikaans, the day will break open (oopbreek), however it is still a vroulike dag ("womanly day"), as in the first stanza. While "female" was used in the first stanza in the English, in the second stanza it is "a womanly day". The English version moves away from the abstract gender designation to a comparison that points towards the mature female body. In the second and third lines, this womanly day is described as "green in greeting/ and squander". "Green" as well as "greeting" creates the impression that the day is still young, an association that is in contrast to the maturity implied in the description "womanly day". "Squander" together with "green" highlights the negative association of inexperience that can be linked to being young. It also adds to it the meaning that the day is wasted.
In the Afrikaans, the day is groen in groete/ en skandering ['green in greeting/ and scanning/ scansion']. The word skandering has two meanings (Odendaal andGouws 2005:1010): it can refer to the electronic scanning of text, graphics or photographic material or it can refer to dividing poetry into metrical feet (scansion). Connotations are therefore made with technology as well as poetry. The new day dawning is an era of technology and electronics where poetry and images are captured electronically rather than in print. However, it is still a day that can be divided into metric feet -a day that can be expressed in poetry. Considering these two meanings, as well as the re-tracing of skandering as "squander", calls forth complex questions about the future. Is the electronic storage of information as well as poetry in vain and therefore a waste of time and energy? Or are new technologies and ways of expression not in vain because they open up new possibilities? Do technology and poetry compete with each other, like the different meanings of skandering, or do they complement each other? How do the two influence each other? These questions are activated in the contact zone of the different meanings of the two versions together, in the interpreting time and place of the in-between, because of the release of the remainder.
The juxtaposition of beauty and violence are again taken up in the last example of re-tracing in the poem. In lines 18 and 19 of the Afrikaans version, an image is painted of a dove coming to nestle in a deserted armoured vehicle in the afternoon of that other, future day described in this poem. Even though the armoured vehicle is abandoned, suggesting that the war has ended, it is still a reminder of more violent times. In the English version, the dove will come to nestle "in the painted war wagon". Pantser [-] from the Afrikaans is re-traced as "painted" and -wa is retraced twice as "war wagon". The reminder of war is still found in the English version but, while it is defused in the Afrikaans by saying that the armoured vehicle is deserted, it is debunked in the English by the word "painted". The word "painted" calls forth connotations of coloured imitations of wood and thin iron railings in playgrounds. As a symbol of peace, the dove comes to show that war is, at least temporarily, something of the past, something far way, part of a different reality. This analysis shows that the two versions not only complement one another but also influence one another. Different meanings result when the form of words in one language is carried over to the version in the other language to open up new meanings. Even though re-tracing is not the only cause for the differences and semantic conflict between the two versions, this translation strategy is a main contributor to the proliferation of difference and semantic conflict between the versions. The phonetic and graphic similarity of re-tracing furthermore links the differences between the versions and the bilingual layout places them in clear contact. As the reader moves back and forth between the versions in the time and space of the in-between, and as similarities and differences are compared, various interpretations can be made based on the two versions together.
The tension of the in-between is not resolved on behalf of the reader. The ambiguities and differences are left as they were so that the remainder of language can be released in all its multiplicity. In 12, this allows the speaker's paradoxical attitude towards the future to be expressed: the speaker is hopeful about the future but also fearful that the future will become yet another day like any other. This paradox is allowed to grow like a rhizome, drawing a map of connections between the two versions to result in a multiplicity of meanings.

Conclusion
"The journey continues and the conversation will carry on," Breytenbach (2009:63) says at the end of the Note in the back of Oorblyfsel/ Voice Over. This article has argued that this is not only the case for Breytenbach's relationship with Mahmoud Darwish and his work, but also for the relationship between the Afrikaans and English versions of Breytenbach's poems. It is a journey and a conversation that readers are invited to become part of, to make their own and to find their way in. Translation, even if it is done differently than in most other cases, plays a key role in establishing this relationship and inviting the reader into this space. Through writing such a collection of poems in this manner, Breytenbach also seems to suggest that there is value in such intercultural and interlingual exchanges, within the world of poetry but also outside of it, just like his friendship with Darwish.
The concepts of the 'contact zone', the 'in-between' and the 'remainder' prove useful in providing a theoretical framework in which the journey and the conversation that is Oorblyfsel/ Voice Over, especially with regard to the nature and role of translation, can be understood and appreciated. Rather than viewing the two versions in isolation, the contact zone allows them to be handled together in a different way than if each was viewed on its own. While the contact zone focuses on the layout and the relationship between the versions, the in-between focuses on the interpretation process that occurs within this relationship. The remainder is focused very specifically on how language allows these interpretations to be made. As the reader moves back and forth in-between the contact zone of the poems, the rhizome of the remainder increasingly grows as more and more connections are made, and more and more meanings are released.