A COMPLEXITY APPROACH TO THE INCIPIENT SIGN SYSTEM OF ZEPHANIAH IN THE 1983 AFRIKAANS BIBLE TRANSLATION by FRANCOIS TERTIUS DE VILLIERS (Student number: 2005023649) DISSERTATION SUBMITTED IN ACCORDANCE WITH THE REQUIREMENTS FOR THE QUALIFICATION MASTER OF ARTS WITH SPECIALISATION IN BIBLE TRANSLATION IN THE FACULTY OF THEOLOGY AND RELIGION UNIVERSITY OF THE FREE STATE BLOEMFONTEIN SOUTH AFRICA DATE SUBMITTED: 14 DECEMBER 2022 JOINT SUPERVISORS PROF JA NAUDÉ AND PROF CL MILLER-NAUDÉ ii DECLARATION I, Francois Tertius de Villiers (student number 2005023649), declare that the thesis hereby submitted for the qualification Master of Arts with specialisation in Bible Translation in the Faculty of the Theology and Religion at the University of the Free State, Bloemfontein, South Africa, is my own independent work and that I have not previously submitted the same for qualification at/in another university/faculty. I also cede the copyright of this dissertation in favour of the University of the Free State, Bloemfontein, South Africa. F. T. de Villiers 2022-12-14 iii ABSTRACT Traditionally a translation is considered to be the result of the interaction between only two elements – the source text and the target text, each defined as being a single text. From the perspective of complexity, the current study argues that the source of a translation should not be viewed as just one text. The current study demonstrates that a translation emerges from a complex source that consists of various elements such as various texts and other factors. These could include other translations apart from the source text that translators consulted. The current study uses the Book of Zephaniah in the Afrikaans Bible Translation of 1983 as an example to detect instances where the translators deviated from their source text. Using principles of Descriptive Translation Studies, the current study compares the translation to its explicitly stated source text (i.e. Biblia Hebraica Stuttgartensia of 1977). Biblia Hebraica Stuttgartensia is a scholarly edition of the Hebrew Bible that contains a critical Hebrew text based on Codex Leningradensis supplemented by a text critical apparatus. Using principles of Descriptive Translation Studies, the translation was compared to its source text to determine the translators’ translation strategies. Describing translation strategies revealed that translators used the strategies of addition, deletion, specification, and transposition. In cases where the Hebrew source text of Biblica Hebraica Stuttgartensia was problematic, the translators sometimes deviated from their source text by following neither the Hebrew text nor the text critical apparatus. In such instances, the translators were forced to base their translation on sources beyond their stated source text; at least some of these additional sources can be identified. This study thus demonstrates the presence of a complex source for the Afrikaans 1983 translation of the Bible in particular, but it also has implications for Bible translations in general. Keywords Afrikaans, Bible translation, Complexity Theory, Zephaniah, textual criticism, translation studies, Descriptive Translation Studies iv ACKNOWLEDGEMENTS It is surely by the grace of God alone, the prayers, love, support, and longsuffering of my wife, Alwin, children, Eileen and John, extended family, and friends both known and unknown to me, that I was able to complete this study. I thank my parents, Francois and Tertia de Villiers, and their parents who instilled within me already as a child the desire to learn and qualify myself to the best of my ability. Special thanks to the University of the Free State, Bloemfontein, South Africa, that assisted me with a partial bursary, as well as my promotors (Professors Jacobus Naudé and Cynthia Miller- Naudé) who guided and enriched me, and often provided me with their personal resources to assist me. I thank Hanlie Rossouw at the Bible Society of South Africa for granting me access to archival materials. I thank both past and current students, Diane Kozelka and Magdalena Lilla Luchici, for providing me with some of the books that I needed for this study at their own expense. I thank Sally Herring for her friendship, prayers, and interest that often helped me to keep going. v TABLE OF CONTENTS List of Abbreviations ……………………………………………………………… vi Chapter 1: Introduction.......................................................................................... 1 1.1 Background. ......................................................................................................... 1 1.2 Hypothesis and Objective ………………………………………………………. 3 1.3 Structure of the Research …………………………………………………………. 3 Chapter 2: Methodology 2.1 Introduction …………………………………………………………………….. 6 2.2 Background of AFR83 …………………………..……………………………... 7 2.3 Shift towards a Descriptive Approach …………………………………………. 14 2.4 A Complexity Approach to a Source Text ……………………………………... 17 2.5 Supplementing DTS with Complexity …………………………………………. 19 2.6 Conclusion ……………………………………………………………………… 20 Chapter 3: Complexity Descriptive Study of AFR83 Zephaniah ………………. 22 3.1 Introduction ……………………………………………………………………… 22 3.2 Zephaniah 1:2 ……………………………………………………………………. 25 3.3 Zephaniah 1:4 ……………………………………………………………………. 27 3.4 Zephaniah 1:7 ……………………………………………………………………. 30 3.5 Zephaniah 1:9 ……………………………………………………………………. 32 3.6 Zephaniah 1:12 …………………………………………………………………… 35 3.7 Zephaniah 1:17a ………………………………………………………………….. 38 3.8 Zephaniah 2:1 …………………………………………………………………….. 39 3.9 Zephaniah 2:2 …………………………………………………………………….. 41 3.10 Zephaniah 2:8d ………………………………………………………………….. 43 3.11 Zephaniah 2:11a …………………………………………………………………. 44 3.12 Zephaniah 2:15 …………………………………………………………………… 47 3.13 Zephaniah 3:15 ……………………………………………………………………. 48 3.14 Zephaniah 3:16 ……………………………………………………………………. 50 3.15 Conclusion ………………………………………………………………………. 51 vi Chapter 4: Conclusions …………………………………………………………… 53 Bibliography ………………………………………………………………………….. 56 Bibles …………………………………………………………………………………… 60 Archival Materials from the Bible Society of South Africa ……...…………………. 61 vii LIST OF ABBREVIATIONS AFR33 Afrikaans Bible Translation, 1933 AFR53 Afrikaans Bible Translation, 1953 AFR83 Afrikaans Bible Translation, 1983 AFR92 Afrikaans Bible Translation, 1983, reprinted in 1992 with reformulations BFBS British and Foreign Bible Society BHK All editions of Biblia Hebraica edited by Rudolf Kittel. BHK1 Biblia Hebraica edited by Rudolf Kittel, 1st Edition, 1906 BHK2 Biblia Hebraica edited by Rudolf Kittel, 2nd Edition, 1913 BHK3 Biblia Hebraica edited by Rudolf Kittel, 3rd Edition, 1937. BHS Biblia Hebraica Stuttgartensia, 1967/ 1977 DRC Dutch Reformed Church / Nederduitse Gereformeerde Kerk DTS Descriptive Translation Studies KJV King James Version, 1611 NBG-51 Nederlandsch Bijbelgenootschap Vertaling, 1951 SV Dutch Sates Authorised Version/ Statenvertaling, 1637 ST Source Text TT Target Text 1 CHAPTER 1 INTRODUCTION 1.1.Background Historically, the 1983 Afrikaans Bible translation (AFR83) was one of a group of translations in more than one part of the world that had the explicit purpose to render what it perceived as its source text (ST) into modern speech. In the South African context of the time, it was a corrective translation and was intended to succeed the Afrikaans Bible translation of 1933 (AFR33), and its revision of 1953 (AFR53). It was corrective in the sense that it was based on newer editions of the Hebrew (i.e., Biblia Hebraica Stuttgartensia / BHS of 1977) and Greek source texts and used a more contemporary idiom than that of AFR33 and AFR53 (Joubert 2020:2). AFR33 and AFR53 were formal word-for-word translations that attempted to reflect the formal structures of the source languages. AFR33 and AFR53 followed the wording of the Bible translation that was familiar to the Afrikaans Bible reading public of the time, which was the Dutch Authorised Version, known as the Statenvertaling of 1637 (SV), which was also a formal word-for-word translation. In contrast, AFR83 explicitly followed the translation theory of functional equivalence, which was regarded as the result of the best scientific research at the time. According to this theory, translation means to reproduce the meaning of the message of the source language text into the target language (Nida & Taber 1969:12). Preserving the meaning of the source language was more important than its formal structures. Translators were instructed to choose the closest natural equivalent in the target language to convey the source language message (Nida & Taber 1969:13). The theory of functional equivalence was prescriptive in nature in the sense that a particular definition of equivalence (Nida & Taber 1969:12; Naudé 2002:47) was used as an instrument of measurement to judge whether equivalence had indeed been achieved or not. From such a prescriptive approach, a translation was judged in emotive terms such as whether the translation was good, bad, correct, accurate, faithful, or not (Hermans 1985:12; Naudé 1999:74). The theory of functional equivalence had the underlying assumption that translation is the result of an interaction between only two elements – the source text (ST) and the target text (TT), each defined as being a single text. According to this view, the task of a translator is merely to communicate this message of the ST across historical, cultural, and linguistic boundaries in a way that is comprehensible for a contemporary audience (Naudé 2022:1). 2 In recent years there has been a shift away from a prescriptive to a descriptive approach to translation because equivalence is both impossible to achieve and poorly defined (Rosa 2010:99, Nord 2016:567). Descriptive Translation Studies (DTS) does not prescribe what equivalence is or what it should be (Hermans 1985:12, Rosa 2010:100). In DTS the fact of a translation implies that equivalence has already been achieved (Hermans 1985:25). Within DTS, equivalence is not a prescriptive, but a descriptive concept (Rosa 2010:99). A prescriptive approach dictates how a translation should be done and a particular definition of equivalence is used like a yardstick to measure whether the desired equivalence has been achieved or not (Hermans 1985:12). Because it is impossible to reproduce in translation an exact copy of the original, the result of such a yardstick test is pre-determined. A translation, when measured against a prescriptive approach to equivalence, is always doomed to fail in conveying the perceived richness of the original and the purpose of the study of a translation is merely to demonstrate its errors (Naudé 2002:47). In contrast to a prescriptive approach to equivalence that dictates how a translation should be done, DTS attempts to describe a translation that has already been done. Since the assumption of DTS is that equivalence has already been achieved, one of the purposes of DTS is to describe what strategies the translators used to achieve that equivalence. In the case of text-specific problems that translators encountered in the ST, DTS can describe what translators did to solve the unique translation problems that they encountered (Nord 2005:168). According to Toury (1985:25) a translation can be viewed as representing the solutions to translation problems. From the perspective of a complexity approach, a reductionist approach to all of the above would assume that AFR83, or any other translation, is the result of the interaction between only two elements. These two elements are the translation’s stated ST and its envisioned TT. According to this view, if a descriptive study of AFR83 is to be done, the task of DTS would be to describe the relationship only between the Old Testament translation in AFR83 and the Hebrew text of BHS with its text-critical apparatus. A complexity approach takes the complexity of reality as its point of departure (Marais 2014:46). According to an approach from the perspective of complexity, a translation is not the result of the interaction between only a single ST and a single TT. Instead, a translation is considered to have emerged from a complex source comprising various incipient texts and other factors. A complexity approach does not attempt to replace existing theories but to supplement them. Supplemented with a perspective of complexity, DTS can attempt to 3 describe not only a translation in relation to its original as if the original were a single text, but it can also describe other influences and texts that were incipient to it. By supplementing DTS with the perspective of complexity, DTS can demonstrate that a translation emerges from a complex source. 1.2.Hypothesis and Objective The hypothesis of the current study is that AFR83 and specifically its translation of the Book of Zephaniah (hereafter called AFR83 Zephaniah) emerged from a complex source or incipient sign system (i.e. the source of AFR83 Zephaniah was not just one other text or just BHS as its explicitly stated ST). The objective of the current study is to demonstrate this hypothesis by identifying and discussing instances where AFR83 deviated from its stated ST, the BHS edition. Since BHS is a scholarly edition of the Hebrew Bible that contains text-critical information that already refers to sources other than the Hebrew text, the definition of a deviation from the ST would be where AFR83 deviated not only from the Hebrew text, but also from the text-critical apparatus. Such instances will demonstrate that for a particular text the translators did not use their stated ST but would by necessity have used other sources as well. This will demonstrate the presence of a complex source that consists of more than just one other text and other influences besides the translation, a complex incipient sign system from which the translation emerged. Another hypothesis is that part of the complex source or incipient sign system would have been other Bible translations that the translators would have consulted. 1.3.Structure of the Research To describe a translation in relationship to its ST places the current study within DTS. The current study seeks to combine the insights of DTS with the theoretical insights of complexity. The dissertation will consist of four chapters. In Chapter 2, I will formulate the theoretical framework and exact procedures of this study in more detail. Chapter 2 will contain four sections. In the first section of Chapter 2, I will give a brief historical background of AFR83. The purpose of the historical section is to indicate that the entire 1983 Afrikaans translation and Zephaniah, in particular, can be positioned within an era in which a prescriptive approach to translation was dominant. The purpose of that 4 section will be to provide the rationale for considering what other elements or texts besides the stated ST might have been part of the incipient sign system of AFR83 Zephaniah. In the second section of Chapter 2, I will describe the shift away from a prescriptive approach to DTS. This will be used to indicate that the current study, in which a translation is compared to its original, even if that original is defined as a complex source, still places the current study within DTS. In a third section of Chapter 2, I will describe some elements of a complexity approach to the source text of a translation and indicate how a complexity approach can supplement DTS. I will pay particular attention to how a complexity approach widens the definition of what is otherwise called the source text. From the perspective of complexity, what is otherwise called the ST should be defined as being part of a larger incipient sign system of which several texts are part amongst other factors. In a fourth section of Chapter 2, I will formulate some principles and methods that can be used to describe AFR83 Zephaniah in comparison to its original. I will draw from some methods of DTS by specifically referring to the description of translation strategies used by the translators. The purpose of that section will be to form hypotheses about which translation strategies would most likely have been used when translators came across text-specific translation problems such as when the meaning of the source text was unclear (Nord 2005:168). In such situations translators would have been forced to look outside of their stated ST to establish their translation and this would demonstrate the translation’s complex source. It is my hypothesis that, in the case of AFR83 Zephaniah, translators would use the strategy of deletion (i.e. an element in the ST is not conveyed in the translation at all), addition (i.e. the TT contains linguistic, cultural, and textual elements that do not appear in the ST at all), transposition (i.e. grammatical structure in ST is conveyed by means of an entirely different grammatical form in the TT), or else a combination of various strategies (Naudé 1999:79). In the same section I will indicate how I intend to combine some elements of DTS with a complexity perspective. According to Marais (2014:15), a complexity approach does not intend to replace reductionist approaches, but to supplement them. I will describe that the specific element of DTS that will be supplemented by a complexity perspective is the definition of the “original” (Lambert & Van Gorp 1985:52-53) that is to be compared to the translation. From the perspective of complexity, the “original” can be considered to be not just one other text, but a complex source consisting of various texts and other factors (Marais 2019:45). 5 Instances where it seems that translators made use of the strategies of addition, deletion, transposition, or a combination of strategies to overcome text-specific problems, could point to the use of other sources besides the translation’s stated source text. Such instances can be compared to other possible incipient texts mentioned in the historical section on AFR83 Zephaniah. If the source of particular readings can be identified as having been imported from other sources, it will not only prove the hypothesis of a complex source for AFR83 Zephaniah as correct, but it will also describe some elements of that complex source even if that is not the main objective of the current study. In Chapter 3, I will do a descriptive study of AFR83 Zephaniah according to the principles and methods described in Chapter 2. I will refer to a selection of texts from AFR83 Zephaniah that seem to deviate from the stated ST. Since the ST of AFR83 was BHS, this would mean that the translators would have had to deviate from the Hebrew text and the text-critical apparatus concurrently. I will compare such deviations to other possible incipient texts. If it can be indicated that, when they deviated from BHS, the translators of AFR83 Zephaniah made use of such other sources, it will not only prove that the hypothesis of a complex source is correct, but it will also describe some elements of AFR83 Zephaniah’s complex incipient sign system. Chapter 4 will be the conclusion of the current study. The conclusion will answer the question if the initial hypotheses of the current study have been proven as correct or not and suggest possibilities for future research. 6 CHAPTER 2 METHODOLOGY 2.1 Introduction As indicated in the previous chapter, the purpose of the current study is to approach the source of the 1983 Afrikaans Bible translation of Zephaniah from the perspective of complexity in order to argue for the existence of a complex source. The translators of AFR83 introduced their translation to the reader by means of an introductory note. In this note they informed the reader about their ST: Die vertaling van die Bybel in Afrikaans wat hiermee aangebied word, is die vrug van vertaalwerk wat in 1970 begin is. Die doel was ’n vertaling wat rekening hou met die ontwikkeling in Afrikaans die afgelope jare en met die resultate van wetenskaplike ondersoek, maar ook so getrou as moontlik aan die grondteks bly; ’n waardige vertaling waardeur Afrikaanssprekendes aangespreek word, buite en binne die kerk, in die erediens en in huislike gebruik, in ons teenwoordige situasie en tyd. As grondteks vir die Ou Testament is die Biblia Hebraica Stuttgartensia (1977) gebruik. Hierdie uitgawe van die Hebreeuse, en gedeeltelik Aramese, grondteks is gebaseer op die oudste volledig behoue manuskrip, die Codex Leningradensis (1008 n.C.). Gegewens uit ouer manuskripte is in die teksgegewens bygewerk. [The translation of the Bible in Afrikaans that is presented here, is the fruit of translation work that began in 1970. The purpose was a translation that takes into account the development of Afrikaans in the past number of years and the results of scientific research, but which is also as faithful as possible to the source text; a worthy translation that may speak to speakers of Afrikaans, inside and outside of the church, for use in the worship service and at home, in our current circumstances and time. As source text for the Old Testament the Biblia Hebraica Stuttgartensia (1977) was used. This edition of the Hebrew, and partially Aramaic, source 7 text is based on the oldest complete preserved manuscript, the Codex Leningradensis (1008 A.D.). Information from older manuscripts has been incorporated into textual information.] The above introductory note informs the reader that the ST for the Old Testament was a scholarly edition of the Hebrew Bible and that the translation was intended as a modern speech translation. From a reductionist point of view, the translation is the result of the interaction between only two factors (i.e. the envisioned AFR83 and BHS). From a complexity perspective, it is the assumption that the translators would have consulted other texts apart from their explicit ST. The purpose of the current study is to argue this point by demonstrating that there are instances that the translators deviated from their ST and must have used other resources as well. The hypothesis of the current study is that the source of AFR83 Zephaniah was not just one other text, but that the translation emerged from an incipient sign system that would have included several texts and other factors. The concept of an incipient sign system includes all influences that contributed to the translation process. These influences can include texts but are not limited to texts only. For the purposes of the current study, I wish to limit the research to consider only possible other incipient texts. Comparing a translation with its original places the current study within the context of Descriptive Translation Studies (DTS). According to Marais (2014:15) a complexity approach is not intended to replace existing approaches that tend to be one-sided or reductionist in nature, but to supplement them. According to Marais et al (2021:119) complexity is not a methodology in the sense that it will provide a researcher with exact procedures. Instead, complexity provides a researcher with an additional framework that will guide his/ her observations. A descriptive study of the relationship between a translation and its original implies a comparison between a translation and usually only one other text that was supposedly its sole source. If the hypothesis of the current study is that the 1983 Afrikaans translation of Zephaniah emerged from a complex source, the purpose of the current study is to detect places where the AFR83 Zephaniah deviated from BHS and to describe the strategies used to establish the translation, with reference to the additional perspective from a complexity approach. Such a description will reveal whether the translators used other sources to establish their translation as well. 8 In the following sections, I provide a brief historical background of the 1983 Afrikaans Bible translation as a whole. The purpose of that section will not be to provide an exhaustive description of its historical and socio-political background, but to illustrate how the AFR83 emerged from an era in which a prescriptive approach to translation was dominant. This will provide the rationale for considering which other texts might have been incipient as part of its complex source. The following section will describe the shift away from a prescriptive to a descriptive approach to translation studies. This will provide the rationale for a method of a descriptive study of AFR83 of Zephaniah that would be able to demonstrate its complex source. Finally, I will describe some principles of a complexity approach and pay particular attention to how a complexity approach can supplement DTS. This will lay the foundation for the procedures that I will follow in Chapter 3, which I believe will demonstrate that AFR83 Zephaniah emerged from a complex source. 2.2 Background of AFR83 In 1970 the three Afrikaans Reformed Churches in South Africa indicated officially that they would support the project of a new Afrikaans Bible translation (Joubert 2020:172) to succeed the Afrikaans Bible translation of 1933 (AFR33) and its revision of 1953 (AFR53), which was in official use in churches at the time. The decision to support a new translation that would succeed it was the result of an historical process stretching as far back as when the first translation of a biblical book into Afrikaans appeared (Naudé 2009:54). Prior to 1933, when the first Afrikaans Bible was introduced successfully to the Afrikaans speaking community, the Bible version long accepted was the Dutch Authorised Version/ Statenvertaling (SV) of 1637. Although the Dutch and Afrikaans languages are similar, by the turn of the 19th century the level of Dutch in the SV was no longer easily understood by speakers of Afrikaans. The first attempts to translate portions of the Bible into Afrikaans were done by S.J. du Toit (Naudé 2009:54). The first translation of a biblical book into Afrikaans was that of Genesis which appeared in 1893. Translations that followed soon after were that of the Gospel of Matthew (1895), the Book of Revelation (1898), Song of Songs (1905), Psalms (1907), Acts, and the Gospel of Mark (1908). These translations were not official in any way, 9 were not intended to replace the SV, and were indeed never fully accepted. Afrikaans speakers continued to use the SV. In 1916 the Orange Free State Synod of the Dutch Reformed Church (DRC) expressed its desire for a Bible in Afrikaans, but the Synod realised that support from the DRC Synods in other provinces, as well as the support of the other two Afrikaans Reformed sister churches (i.e. Nederduitch Hervormde Kerk van Afrika and Gereformeerde Kerke in Suid-Afrika) was necessary. In 1917 the British and Foreign Bible Society (BFBS) offered to support the translation project and its publication financially. This support was offered on condition that the final product should resemble the wording of the Dutch SV. The SV was to be the source text and a translation into Afrikaans was merely to be checked against the Hebrew and Greek ST (Naudé 2009:55). This trial translation of the four New Testament Gospels and the Psalms was published in 1922 but it was rejected by the Afrikaans Bible reading public. The result was the decision to embark upon a full translation project from the Hebrew and Greek ST. For the Old Testament the ST was to be Rudolf Kittel’s 2nd scholarly edition of Biblia Hebraica, published in 1913 (BHK2). This edition was based on the Second Rabbinic Bible published in 1524-1525 in Venice by Daniel Bomberg supplemented by a text critical apparatus. The actual Second Rabbinic Bible, which contained the Hebrew text supplemented with Rabbinic commentary, was the source text of both the SV and King James Version of 1611 (KJV). The translators of the 1933 Afrikaans translation followed as far as possible the wording of the SV. This included using the same source texts as the SV. For the New Testament the translators of the 1933 Afrikaans Bible used the Greek New Testament prepared by Desiderius Erasmus in 1516 despite having access to later and better editions of the Greek New Testament by then. They did this because they believed that it was the source text of the SV translation of the New Testament (although it was not, according to Naudé 2011:9). According to Naudé (2009:62) the translators of the 1933 Afrikaans Bible used the translation method of following the ST word-for-word. This translation method was followed by the SV as well as the KJV and was regarded as normative (Naudé 2009:63) at the time. The revision of the 1933 translation published in 1953 was corrective in the sense that some changes in spelling and word choice were incorporated (Naudé 2009:67). By then Rudolf Kittel’s Biblia Hebraica 3rd Scholarly Edition (BHK3) was published in 1937 as well as its successor of 1951, which contained information from the Dead Sea Scrolls, which had been recently discovered in 1947. The significance of BHK3 was that it was based on a different source text than the second edition. The Codex Leningradensis, as opposed to the Second Rabbinic Bible, was the 10 Hebrew source text of BHK3 and of subsequent editions. The Second Rabbinic Bible contained nine textual differences compared to Codex Leningradensis and therefore also editions of Biblia Hebraica after 1937 (i.e., 1 Kings 20:38; Proverbs 8:16; Isaiah 10:16; Isaiah 27:2; Isaiah 38:14; Jeremiah 34:1; Ezekiel 30:18; Zephaniah 3:15; Malachi 1:12). Relevant to the study of Zephaniah, it is clear from the difference between the rendering of Zephaniah 3:15 in AFR33 and in AFR53 that the revisors of AFR53 indeed did have BHK3 available to them. AFR33 reads Zephaniah 3:15 as jy sal geen kwaad meer aanskou nie (“you will see no more evil”). This reading assumes י רְאִּ which appears in the (ראה you will see” based on the root“) תִּ Second Rabbinic Bible and Kittel’s second scholarly edition of the Hebrew Bible. Zephaniah 3:15 in AFR53 reads jy sal geen kwaad meer vrees nie (“you will fear no more evil”), which assumes the reading י ירְאִּ ,which is the reading of Codex Leningradensis ,ירא based on the root תִּ BHK3, and subsequent editions. It is interesting that the translation of the Dutch Bible Society of 1951 (NBG-51) also followed this reading. Shortly after the publication of the revision of 1953, a new commission was formed in 1958 tasked with a further revision (Joubert 2020:170). The commission did not function for long since it became clear at an early stage that the changes in the everyday use of Afrikaans, as a new and fast-developing language, meant that AFR53 no longer represented the language use of contemporary Afrikaans speakers. Either a full revision or a fresh translation was necessary. In 1960 Prof. P.F.D. Weiss visited the BFBS in London to request a full revision. The request was granted after some reluctance (Joubert 2020:170). In 1961 a committee was formed with the revision in mind. By 1966 the committee ground to a halt. By then only Prof. B.B. Keet was left as one of its original members, and he vacated his position in favour of Prof. E.P. Groenewald. A turn towards making a fresh translation came in 1967 when the Bible Society of South Africa organised a translation seminar at Turfloop. Eugene A. Nida, the secretary for translation at the American Bible Society, as well as the person who developed the theory of functional equivalence (Nida & Taber 1969), was invited as the guest speaker. The Secretary of the BFBS, Rev. J.T.M. Arkel was also present. Based on presentations at the seminar regarding difficulties in translating the Old and New Testament, as well as the latest developments in translation theory, Rev. J.T.M. Arkel posed the question whether an entirely new translation into Afrikaans should be pursued rather than another revision (Joubert 2020:170). Nida responded favourably to this proposal. The principles adopted for what would become the 1983 Afrikaans Bible (AFR83) were formulated at that seminar. The main characteristic of 11 the new translation would be that the meaning of the original text should be transferred in a language that is comprehensible for the lay person and the younger section of the population, and that it should be faithful to the source texts (Joubert 2020:171). The first meeting of the Translation Commission of what would become AFR83 was held on 2nd of February 1971. The minutes of that meeting contain information that confirm that the 1983 Afrikaans Bible translation emerges from an era of a prescriptive approach to equivalence and translation. In the minutes, there is reference to the translation seminar of 5 July 1968 mentioned above. The Commission stated that they intended to produce a translation into Afrikaans that would follow the principles discussed at the mentioned seminar as opposed to a translation that adheres to the form of the ST such as the AFR53 translation. There is reference to traditional church and technical theological terms that should be avoided and simplified. These include terms such as geregtigheid (“justice”), goedertierenheid (“lovingkindness”), and geregtigheid deur geloof (“justification by faith”). The Commission decided that sentences should be shorter, simplified, and what is implicit in the ST should be made explicit in the TT. The minutes of the meeting explicitly mention the New English Bible (NEB) of 1970. The NEB did not follow the theory of functional equivalence (i.e. the theory adopted by the translators of AFR83), but it did have a similar objective as AFR83 in that it was intended to be a translation into modern speech. The meeting discussed in detail some of what the Commission members felt were errors from the side of the NEB translators. One of these was that the NEB used formal “thee” and “thou” language. According to the minutes of the meeting on 2nd February 1971, the Commission decided to use BHK3 as ST for the Old Testament since it was the latest critical edition of the Hebrew Bible available at that stage. This is significant, because it means that the ST of the 1983 Afrikaans Bible would be based on the Codex Leningradensis, a different Hebrew text from that used by the 1933 translation. Some other decisions include the Commission’s manner of operations, names of representatives of the three Afrikaans Reformed Churches, as well as the allocation of translators to specific biblical books. The minutes of the Translation Commission meeting of 2nd February 1971, are very significant for the current investigation. One of the main hypotheses of the current study is that not only AFR83’s explicitly stated source text, but also other texts would have been incipient to it. Since the meeting makes mention of various texts and concepts, it is possible to consider that these were also incipient elements that formed part of AFR83. 12 The Commission decided that BHK3 would be their source text for the Old Testament. This is significant because, when the first edition of AFR83 was published, the introductory note to the reader stated that a different text served as source for the Old Testament by then. According to this introductory note, the source text of the translation was not BHK3, but its successor, Biblia Hebraica Stuttgartensia (BHS) of 1977. It surely makes sense that, as a later edition of Biblia Hebraica would become available, that the translators would use it and that is why BHK3 was no longer used after 1977. It does, however, also imply that for almost half the time that AFR83 was in the making, translators were using the predecessor of BHS and its text- critical apparatus. It would be worthwhile to consider how some elements of page formatting, contents, or presentation of text-critical notes in the two different publications’ apparatus might have influenced the translators. The Commission decided explicitly to accept the theory of functional equivalence as normative and took further decisions concerning their method of translation and operation. These decisions, influenced by their accepted theory of functional equivalence, included their choice of ST, their vision of what their end-product should look like, and what processes of feedback and cross-checking they would follow to make sure that their translation measured up to their conception of equivalence. Their decision to produce a more “dynamic” translation means that they must have had some conception of how they did not want their translation to look. It is possible that the translators would have consulted some formally equivalent translations such as the SV, AFR33, and AFR53, be it as a source of exegesis, text-critical information, or as an example of the kind of translation that they would attempt to move away from. The Commission mentions the New English Bible (NEB) of 1970 by name. Although the Commission proceeded to discuss some elements of the NEB that they did not agree with, the implication is that by the time the translation of the AFR83 was in progress, the NEB would already have formed part of AFR83’s incipient complex source and it might already have had an influence on the translators, because they would have consulted it. It also opens the possibility that other translations that had a similar objective as AFR83 would have had an influence upon it as well, such as the Good News Bible (GNB) (Naudé 2005:82). For the purpose of the current study, the name of Prof. P.A. Verhoef in the minutes of 2nd February 1971 is significant for several reasons. First, the Commission decided that he would be part of the committee that would be responsible for the translation of the Old Testament and that he would be responsible for the Book of Zephaniah, amongst others. 13 Second, the mention of his name is also important because it implies yet another possible incipient text to AFR83. Prof P.A. Verhoef was the promotor for a Master of Theology thesis on Zephaniah by J.P.J. Olivier at the University of Stellenbosch. Olivier (1973:iv) states explicitly that his thesis is intended as a help for the translators of the new Afrikaans translation that was underway at that stage. Olivier’s thesis attempts to make his translation according to the theory of functional equivalence and records meticulously how he followed the procedures prescribed by Nida & Taber (1969) and he did extensive research concerning text-critical problems in the Book of Zephaniah. The above brief historical overview requires a few brief summarising statements to establish a logical connection between the current section of this chapter and the one that follows. These summarising statements concern ultimately which texts might have formed part of AFR83 Zephaniah’s complex incipient sign system and are worth comparing to AFR83. The NEB is mentioned by name in the meeting of 2nd February 1971. This means that they would have consulted it as an example of a modern speech translation. The Commission stated explicitly that they wanted to move away from formally equivalent translations of which the KJV, AFR33, AFR53, and NBG-51 were examples. Since the translators indicate specifically that they wanted to move away from such formally equivalent translations, this would mean that they would have consulted them even if they were examples of how they did not want their translation to look, or that they would have consulted them as a source of exegesis and text- critical information before formulating their own “dynamic” translation. Their decision for which ST to use for the Old Testament is significant, because it means that not only BHS as explicit source text, but also BHK3 would have had at least some influence since the translators would have used it for at least 6 years before AFR83 was published. The mention of Prof. P.A. Verhoef implies that the M.Th. thesis of J.P.J. Olivier would have been taken into account as well; because Prof. P.A. Verhoef, as the one responsible for the translation of Zephaniah would have had to read it by 1973 when J.P.J. Olivier completed his thesis. The Translation Commission decided explicitly that it wanted to follow the theory of translation as formulated by Eugene A. Nida, with the result that the GNB would have had an influence when it became available in 1976 as well. This is because the translators of the GNB, just like the AFR83, desired to follow the theory as formulated by Eugene A. Nida. The decision by the AFR83 translators that this theory of equivalence (Nida & Taber 1969) would be normative to them, implies that the Commission held a prescriptive view of translation. In the following section 14 of this chapter, I give a brief description of the move away from a prescriptive to a descriptive approach. 2.3 Shift towards a Descriptive Approach In the previous section I indicated that the 1983 Afrikaans Bible translation finds itself in an historical context in which a prescriptive approach to translation and equivalence was dominant. The movement away from a prescriptive to a descriptive approach to translation in recent years can be attributed to the realisation that equivalence is poorly defined (Nord 2016:567) and, even if it were to be defined more clearly, equivalence is impossible to achieve. To illustrate this, Nord (2016:570) uses the example of the translation of ancient miracle stories. An approach that attempts to achieve equivalence would have to determine what the meaning of the ST was for its original audience and how they would have responded (Nida & Taber 1969:23). Based on that, a translator would have to achieve equivalence by reproducing the meaning of the ST in such a way that it would, if not generating the same effect, at least mean the same to the target audience as it did for the original audience. According to Nord (2016:570) “it would not make any sense to aim at recreating the functions or effects the original texts had or may have been intended to have for their receivers” because the modern context is just so far removed from the original audience that it is not possible for a modern audience to listen to a miracle story with the same expectations as the original ancient audience. According to Nord (2005:25), the concept of equivalence has been debated since its inception. The aspiration for equivalence gives rise to untenable contradictions (Naudé 1999:74). The reasons why a translation might fail in the ideal of achieving equivalence vary – a translation may be too literal or not literal enough. Measured against the yardstick of equivalence, a translation that follows its original too closely, too faithfully, too literally, is deemed as equally lacking in equivalence as a translation that is not faithful enough or too free. Despite various attempts to define the concept of equivalence and on what levels a translation should be deemed as equivalent, one of the problems in using equivalence as a requirement for what translators should be doing, is the “inherent fuzziness” of the concept (Nord 2016:567). According to Hermans (1985:7), if a prescriptive concept of equivalence is used to judge a translation, translation would be a futile exercise by default. This is because a translation, when compared to the perceived richness of its original, would always be found to be wanting. According to Hermans (1985:8), there are those who consider translations not only as “second-hand” but 15 also as “second-rate”, as an inferior object of study. Such an opinion results that even when a translation is studied, the object of study ultimately remains the ST – the study of a translation would merely be to demonstrate its errors when compared to its original. Such an approach takes the original to be normative, the forever unattainable ideal which a translation can never achieve since it is impossible to produce a translation that is the exact copy of its original (Naudé 1999:74). According to Nord (2005:26) various attempts to define equivalence have not achieved a clearer definition. What has been revealed is rather that the concept is ambiguous. Because equivalence is poorly defined, there has been in recent years a shift away from a prescriptive to a descriptive approach to translation. The shift has been away from defining principles that prescribe how a translation should be done, to a description of how a translation has actually been done. Descriptive Translation Studies (DTS) analyses existing translations by asking the question, “Why is there what there is?” (Rosa 2010:94). Holmes (1988:72) considers DTS as an attempt to describe, amongst other things, what happens in the “little black box” of the translator’s mind when creating a “more or less matching text” in another language. According to Holmes (1988:72) there are three objects of study within DTS: translation as product, function, or process. A product oriented DTS examines the translation as a finalised product that describes the translated text itself or different translations of the same original. Function oriented DTS examines what function the translation has in society (e.g., in comparison or in contrast with the translation’s intended function). A process oriented DTS describes the translation process and could focus on internal or external influences upon the translator. According to Hermans (1985:10) DTS arises from the need not for theories about equivalence from which procedures may be prescribed, but rather DTS arises from the need for practical case studies to describe and account for what translators actually did to establish their translation (Hermans 1985:13). According to Toury (1985:18) translations are phenomena of reality that can be observed. They attest to how translators went about solving translational problems (Toury 1985:21). One of the objectives of DTS is to reconstruct the process that gave rise to practical translation decisions. This reconstruction could entail the description of the norms, constraints, and strategies that determined translation decisions (Toury 1985:22). Lambert & Van Gorp (1985:52-53) provide some practical guidelines for a descriptive study consisting of four steps to describe the relationship of translated texts to their source texts. The 16 first step is to collect preliminary data such as the title page, introductions, footnotes, and information regarding the strategy of translation. The second step is to collect data on macro- level (e.g., how titles, chapters, and paragraphs are divided). The third step is to collect micro- level data. It is on this level that a selected text is considered. Under investigation could be a selection of words, dominant grammatical patterns, formal literary features such as meter and rhyme, modality of verbs (e.g. passive or active, subjunctive or indicative), and language levels (e.g. formal, informal, or popular). It is during this third step that translation strategies of specific texts can be described (Naudé 1999:79). The fourth step is to collect data from the translation’s system in which it functions. According to Lambert and Van Gorp (1985:49) no study can possibly be exhaustive. Researchers will need to select what specific elements of a translation will be described. From the above, it can be inferred that the main difference between a prescriptive approach, such as the one followed by AFR83, and a descriptive approach is the definition of equivalence. According to Rosa (2010:97), within DTS, equivalence is a descriptive concept. The observable fact of a translation assumes that equivalence has already been achieved and the task of DTS is to describe what translators did to achieve that. The fact of the translation assumes that translators encountered certain translation problems and their translation represents their solutions. In order to establish a logical connection from the current to the next section, a few summarising comments are necessary in order to indicate how a study with DTS can be reductionist in nature from the perspective of complexity, but also at which point DTS can be supplemented by a perspective of complexity. In this section I indicated that a descriptive study of a translation can be conducted at various levels. DTS is the description of the relationship between a translation and its original. One of the ways to describe the relationship between a translation and its original is to consider the use of translation strategies. The assumption is that the translation represents practical examples of what translators did to overcome translation problems. It would seem that, implicit to this view (i.e. DTS is the description of the relationship between a translation and its original) is the assumption that what is to be described in DTS is the relationship between a single source text and a single target text. It is at this point that a complexity approach can supplement DTS. 17 2.4 A Complexity Approach to a Source Text Earlier in this chapter, I indicated that the 1983 Afrikaans Bible translation emerged from an era in which a prescriptive approach to translation and equivalence was dominant. I briefly described the shift away from a prescriptive to a descriptive approach to translation. If a particular text is considered to be a translation of another, equivalence is the relationship between them, and it is assumed to have been achieved already. The task of DTS is to describe and explain that equivalence (Rosa 2010:100). From the above it can be inferred that the implicit assumption of both a prescriptive and a descriptive approach, is that a translation is the result of the interaction between only two elements defined usually as a single ST and a single TT. ST and TT are considered each to be texts only, that they are fixed, static, and mutually exclusive (Marais 2014:99). From the perspective of complexity, this is a reductionist view. A reductionist approach tends to break down and separate wholes into its constituent parts (Larsen-Freeman 2013:269). Within a paradigm of simplicity or reductionism, each small part is studied in isolation. Reductionism has the tendency to separate and contrast binary relationships, and contradictions are seen as error (Larsen-Freeman 2013:370). A complexity approach takes the complexity of reality and its interrelatedness as its point of departure (Marais 2014:26). From a complexity perspective, reality is emergent. Reality, such as a Bible translation as an element of reality, is the higher-level result of an interaction of various lower-level incipient elements. Applied to translation, Marais (2014:10) argues that a translation, as a phenomenon, as part of reality, is a new whole that has emerged through the input of various meaning-bearing signs of different types. This means that a translation does not emerge from the interaction between only an ST and a TT, each understood to be a single text, nor solely from the interaction of texts, but from various elements that were incipient to it. When approached from the perspective of complexity, what is usually called the ST and TT should be defined in a non-reductionist way. Marais (2019:44) does this by redefining what is otherwise called the ST and TT as an incipient text and subsequent text, respectively each being part of a larger incipient and subsequent sign system of which texts are just one element. According to Marais (2019:44) a translation as product and process emerges from a variety of cause-and-effect relationships that gave rise to it. A translation is the “much coming from little” (Marais 2014:48), the new higher-level whole emerging from the interaction of parts at 18 a lower level. The incipient sign system from which a translation emerged, includes various texts that might have influenced it, as well as non-textual elements such as historical circumstances, the context of the translators themselves, the belief system of the translators, and other meaning-bearing elements that gave input to the translation process (Marais 2019:45). According to Marais (2019:46) subsequent meanings, texts, or other elements that arise from the translation process, themselves feed into new incipient processes. As an example, Marais (2019:46) specifically mentions how a particular subsequent text, such as a Bible translation’s interpretation of a specific text, can be incipient to influence the interpretations of other translations in future to form part of the complex source from which a new translation will emerge. A complexity approach assumes that the translation is a new whole that emerges. Emergent reality is not merely the result or the sum-total of its parts. From a perspective of complexity, the value of the whole that emerges is more, is different, and has “acquired qualitatively new properties” than the sum-total of its parts (Marais 2014:49). Complexity assumes that reality has a high level of inter-relatedness (Marais 2014:24). Complexity takes a meta-stance in which the tension between binary relationships is not resolved by making a choice for one or the other, but the assumption is that they are both part of reality and that their value lies in their relationship with each other (Marais 2014:15). From a complexity perspective, the value of binaries is that they are both part of the various cause-and-effect relationships from which reality emerges (Marais 2019:43). A binary usually upheld in translation studies is that of source and target. From the perspective of reductionism, both are seen as fixed, static, and mutually exclusive. From a reductionist point of view, the task of a translator is to communicate the fixed and static meaning of usually a single ST across cultural and linguistic boundaries (Naudé 2022:1) in a single TT of which the meaning is assumed to be equally fixed and static. A complexity approach does not make a choice between source and target in this way (Marais & Meylaerts 2019:6) and it presupposes that neither source nor target is stable and static (Marais 2019:45). From the meta-stance of a complexity approach, the binary of source and target do not need to be separated from each other. They are in constant and creative tension, being part of the translation process concurrently. Within a complexity approach, both source and target can be accommodated at the same time. From a complexity perspective, reality is not only made up 19 of things, but also of the relationship between things (Marais 2014:25). According to Marais (2019:49) meaning is not frozen within a source but meaning and translation emerge from the complex interaction between the totality of elements in the incipient system. According to Marais et al (2021:119) complexity is not a methodology in the sense that it will provide a researcher with exact procedures. Complexity provides a researcher with an additional framework that will guide his/ her observations. A complexity approach does not attempt to replace existing theories but to supplement them. In order to establish a logical connection between the current section and the following, a few summarising statements are necessary. I have indicated that a complexity approach takes a stance against reductionism and that it takes the complexity and interrelatedness of reality as its point of departure. A complexity approach does not intend to replace existing theories, however reductionist they might be. Instead, its purpose is to supplement these theories and methods by another perspective. I have indicated that a descriptive study compares a translation with its original. A reductionist approach to DTS would assume that what is traditionally considered to be the source text is a single text other than the translation. In the section below I will indicate how this element of DTS can be supplemented by a perspective of complexity. 2.5. Supplementing DTS with Complexity A complexity approach does not negate the methods of DTS such as that described by Lambert & Van Gorp (1985:52-53) and the description of translation strategies (Naudé 1999:79). A perspective of complexity adds an additional perspective to it. A central concept in a complexity approach is emergence (Marais 2019:47). A DTS that attempts to describe translation strategies with the perspective of complexity, should proceed from the hypothesis that the translation emerged from a complex source that consisted of more than one text. In order to conduct a descriptive study from the perspective of complexity, a descriptive study of translation strategies could proceed initially in the way that a reductionist approach usually would. This could be done by comparing only the explicitly stated ST to the translation. One of the ways to compare a translation to its original is to describe the translation strategies used by the translators. This would be a description on the micro-level according to the method 20 developed by Lambert and Van Gorp (1985:52-53). Such a description should be able to detect where translators deviated from their stated ST. At this point DTS can be supplemented by the insight of complexity that a translation emerges from a complex source that could not have consisted of just one other text. Such deviations from the explicitly stated ST can then be considered as examples of where translators would have used other sources to establish their translation. DTS can be supplemented by the additional question, “If the translators did not use their stated ST in this instance, what did they actually use?” If this question could reveal the use of a source other than the stated source text, it would not only demonstrate that the translation emerged from a complex source that consisted of at least more than one text, but it could also describe some elements of that complex source. Applied to AFR83 Zephaniah, one of the ways to demonstrate its complex source could be to do a descriptive study according to the methods described by Lambert & Van Gorp (1985:52- 53) on the micro-level, but supplement it by the insights of complexity. Such a study could proceed by using a line-for-line or phrase-for-phrase comparison between the stated ST (i.e. BHS) and AFR83 Zephaniah. The description of the translators’ strategies will reveal what the translators did to establish their translation. This should also reveal where they deviated from their ST because they came across text-specific problems. Such text-specific problems would include that the source text is not readily comprehensible (Nord 2005:168). It is my hypothesis that when translators came across text-specific problems where the source text is difficult to understand they would have been more inclined to use some specific strategies more than others, such as deletion (i.e. an element in the ST is not conveyed in the translation at all), addition (i.e. the TT contains linguistic, cultural, and textual elements that do not appear in the ST at all), transposition (i.e. grammatical structure in ST is conveyed by means of an entirely different grammatical form in the TT), or else a combination of various strategies (Naudé 1999:79). Another hypothesis is that, when confronted with such text-specific problems, the translators would have consulted other texts, not only their explicitly stated ST (Reid 2021:45). 2.6. Conclusion In the current chapter I described that the historical era from which the Afrikaans Bible translation of 1983 emerged was one in which a prescriptive approach to translation was dominant. I indicated that I believe a study within DTS, supplemented by insights from a complexity approach, could be used to compare the translation of Zephaniah with its explicitly 21 stated ST and then identify instances where AFR83 deviates from BHS. A description of the strategies used at such instances, could reveal that translators used sources other than the explicit ST for their translation. Since BHS was the explicitly stated ST for AFR83 as a scholarly edition that contained both the Hebrew text and a text-critical apparatus, DTS from the perspective of complexity would have to identify instances where the translators deviated not only from the Hebrew text, but also from suggestions in the text-critical apparatus. This would imply that translators used sources other than their ST to establish their translation and this would prove the main hypothesis of the current study as correct. If a description of some translation strategies reveal that the translators deviated from their ST, it implies that they would have used some other source to establish their translation, even if that source cannot be identified. If the comparison between various possible incipient texts could reveal at least some possible sources that the translators of AFR83 Zephaniah might have used, it could describe some elements of that complex source that was part of AFR83 Zephaniah’s incipient sign system. In the next chapter I will conduct the descriptive study of a selection of texts from the 1983 Afrikaans Bible translation of the Book of Zephaniah from a perspective of complexity. If the comparison of the relationship between AFR83 and its ST reveals influences apart from the translation’s explicit source text, it will demonstrate the translation’s emergent nature and therefore also its complex source. This would prove the hypothesis of the current study that the 1983 Afrikaans Bible translation of Zephaniah emerged from a complex source as correct. Inadvertently it might also describe elements of that complex source. 22 CHAPTER 3 COMPLEXITY DESCRIPTIVE STUDY OF AFR83 ZEPHANIAH 3.1. Introduction The translators of the Afrikaans Bible of 1983 (AFR83) informed their readers in the introductory note that the Hebrew text of Codex Leningradensis as reproduced in Biblia Hebraica Stuttgartensia (BHS) and its text-critical apparatus served as their source text (ST). This reflects the traditional notion that a translation is the result of the interaction between only two elements (i.e. a source text and a target text) (Naudé 2022:1), each understood to be a single text. From the perspective of complexity, this is a reductionist view. The source of a translation cannot be just one other text. This is because a translation, as an element of reality, is considered to have emerged from the interaction of various elements of an incipient sign system. Such a sign system of various incipient elements would have included other texts, but also other factors. Such an incipient sign system would include other texts and even non- textual factors such as constraints upon the translators, their historical context, their socio- political context, their belief system (Joubert 2020:11), and whatever other elements might have impacted upon the translators (Naudé 2022:1). From the perspective of complexity, a translation, as an observable element of reality (Toury 1985:19), is considered to have emerged from a complex interaction of various relationships within a complex system, a complex system that would have contained various factors (Marais & Meylaerts 2019:5). This means that, included in this system, are indeed texts, but the incipient sign system does not contain only texts. There would have been various meaning- bearing factors, various elements of different kinds that would have contributed to the translation process. This is what Marais (2019:8) calls an incipient sign system as opposed to referring only to a source text. Since an incipient sign system contains various texts and other factors – texts that are part this system should be called incipient texts and texts that emerge from the translation process should be called subsequent texts (Naudé 2020:1), always bearing in mind that texts are but one element of an incipient sign system. A translation, as a subsequent text, is itself part of new incipient processes from which yet other subsequent translations or texts will emerge (Marais 2019:46). 23 The above makes it clear that the claim by the translators of AFR83 that their source was just one other text, is reductionist in nature and does not take the complex and emergent nature of translation into account. As indicated earlier in this thesis, the purpose of the current study is to test the hypothesis that the source of AFR83 Zephaniah was not just one other text, as is claimed by the translators, but that AFR83 Zephaniah emerged from a complex incipient sign system that would have consisted of other elements and other texts as well. This means that more texts than just its stated ST would have formed part of AFR83’s complex source. To test this hypothesis, it will be necessary to demonstrate instances where the translators of AFR83 Zephaniah must have made use of a source other than BHS to establish their translation. These would be at instances that the translators deviated from their stated ST. Since the translators of AFR83 claimed that their ST was BHS (i.e. a scholarly edition that includes possible textual variants) it must be clearly defined what should be considered to be a deviation from the ST. Since the source text claimed by the translators of AFR83 contained both the Hebrew Masoretic text accompanied by a text-critical apparatus, a deviation from the ST should be considered as instances where the translators deviated not only from the Hebrew text, but also from the apparatus. This will demonstrate that a particular text’s translation must have come from a source other than the stated ST. This will demonstrate that AFR83 Zephaniah emerged from a complex source that consisted of at least more than one other text. I must state explicitly that I realise that the concept of an incipient sign system encompasses all meaning-bearing factors that could have given an input to the translation process and that texts are but one part of this (Marais 2019:45). However, for the purpose of the current study, I will limit myself to detecting the presence of other possible incipient texts. If there are instances in which other incipient elements become evident, these will be noted as well, but they are not the primary purpose of the current study. I will proceed by selecting texts from the Book of Zephaniah. Using some principles of Descriptive Translation Studies (DTS) (Lambert & Van Gorp 1985:52-53) I will compare the Hebrew text and the text-critical apparatus of BHS to the translation. Working on the micro- level (Lambert & Van Gorp 1985:53), I will describe the translation strategies used to establish the translation (Naudé 1999:79). This should reveal instances where the translators deviated from both the Hebrew text and the text-critical apparatus of BHS. I hypothesise that this could have been done especially by the translation strategy of deletion (i.e. instances where translators did not convey an element of the source text into the target text at all), addition (i.e. 24 the translation contains elements that are not contained in the source text at all), and transposition (i.e. a strategy that involves changing the grammar of the source text by conveying it by means of a different grammatical structure in the translation), or a combination of other strategies. Earlier in this dissertation I argued what other texts might also have been incipient to AFR83 Zephaniah (i.e. KJV, SV, NBG-51, AFR33, AFR53, NEB, GNB, BHK3, and Olivier’s 1973 M.Th. thesis). Instances where AFR83 Zephaniah deviated from both the Hebrew text and the text-critical apparatus of BHS will be compared to these possible incipient texts. From the perspective of complexity, there must have been other incipient texts or other elements besides the ones mentioned above. Due to the sheer vastness of the topic and because it is humanly impossible for any human being to be fully aware of absolutely everything around them (Marais 2019:47), I will pay attention only to the above mentioned possible incipient texts (Marais 2019:47). The reason is that, although a comparison between AFR83 and other incipient texts might even reveal the actual source other than BHS used by translators in specific cases, the ultimate purpose of this study is not to identify those other sources, but merely to demonstrate the fact of a complex incipient system. However, if some of the above texts could be demonstrated as having been incipient in that they had an influence on the translators, it will justify the claim for the presence of a complex source even more. This chapter is divided into sections as follows. The title of each section will be the verse reference of the text from the Book of Zephaniah under consideration. The Hebrew text of BHS (1977) will be quoted followed by AFR83 for that verse with an English back translation. Then I will present the various other possible incipient texts in chronological order of their publication. Since KJV represents a more word-for-word tradition of translation, I will quote the KJV as well for the sake of having a literal English translation of the Hebrew at hand. I will discuss the text-critical apparatus and translation strategies used by the translators of AFR83 Zephaniah. I will discuss the possible influences that might have guided the translators of AFR83 Zephaniah in each verse. The conclusion of each section will be a determination of whether the translators deviated from their ST (i.e. the translation follows neither the Hebrew text, nor the text-critical apparatus) or not. If so, then it can be inferred that the translators must have used another source to establish their translation and this will demonstrate the presence of a complex source, a complex incipient sign system of which more than one text must have formed a part. 25 3.2. Zephaniah 1:2 ה׃ ה נְאֻם־יְהוָָֽ ל פְנֵֵ֥י הָאֲדָמָָ֖ ַ֛ ל מֵע ף כ ֹּ֗ ף אָסֵֵ֜ אָס ֹ֨ AFR83: Ek gaan ’n einde maak aan alles op aarde, sê die Here. [I am going to make an end to everything on earth, says the Lord.] KJV: I will utterly consume all things from off the land, saith the LORD. SV: Ik zal ganselijk alles wegrapen uit dit land, spreekt de HEERE. AFR33: Saamskraap, wegraap sal Ek alles van die aarde af weg, spreek die Here. NBG-51:Volkomen zal Ik alles van de aardbodem wegvagen, luidt het woord des HEREN. AFR53: Saamskraap, wegraap sal Ek alles van die aardbodem af weg, spreek die Here. NEB: I will sweep the earth clean of all that is on it, says the Lord. GNB: The Lord said, “I am going to destroy everything on earth.” AFR83 translates this verse as a simple future tense. The translation reflects only one verb. It is not clear how AFR83 understood the grammar of this very problematic verse. The combination of the Qal infinite absolute ף in combination with the Hiphil אסף of the root אָס imperfective first person common singular אָסֵף understood as a volitive (i.e. “let me cause to come to an end”) of the root סוף is problematic in the Hebrew text. Text-critically there is no evidence to emend the consonantal text. Ancient versions reflect different vocalisations of what must have been the same consonantal text. The meaning of the text-critical note in BHS for this verse is not entirely clear. BHS suggests that אָסֵף should be vocalised as סֵף סֵף What the analysis of .א ,is supposed to be, is not clear א because even this vowel pointing can have more than one possible meaning. BHK3, which was the source of AFR83 Zephaniah initially before the publication of BHS in 1977, contained a text-critical note that suggested that the reading should be emended to read ף ס ף אֶאְֶ thus creating a Qal imperfect first person common singular in combination with a Qal אָס infinite absolute of the same root אסף. This would create a tautological infinitive construction that could be translated as I will surely gather or I will utterly take away. This reading has its origin in the original publication of the Second Rabbinic Bible upon which BHK1 and BHK2 were based. The original publication of the Second Rabbinic Bible edited by Jacob Ben 26 Chayyim and published by Daniel Bomberg in Venice in 1524-1525, contained commentary of Rashi in the margin of this Hebrew text. According to that commentary the reading represents a misspelling and should be understood as ף ס ף אֶאְֶ BHK1 and BHK2 presented .אָס this commentary in their text-critical apparatus. This text-critical suggestion was contained in BHK3 as well. It seems that this was the understanding of both the KJV and SV which translated this text as a Qal imperfect first person common singular in combination with a Qal infinitive absolute of the root אסף, thus creating a tautological infinitive which it translated as I will utterly destroy. BHS itself presents a briefer text-critical note that proposes that אָסֵף should be emended to סֵף א but without explanation. It is not clear if the intention is that this verb should be understood as a Qal participle with the sense of imminent action which could be translated as I am surely about to gather/ take away. It is not clear if this is the reading adopted by AFR83 since it translated this expression as a simple future tense using only one verb. Another possibility of how AFR83 understood this expression is that a letter aleph was omitted by means of haplography or that it is due to the same phenomenon that appears in the Qal Imperfect first person common singular of the verb root אמר. The verb for “I will say” appears as ר מ in the א Hebrew Bible. In such a case the assumption is that the visible aleph is the prefix of the Imperfect first person common singular while the root aleph has fallen away due to its function as a silent letter. Whether read with a different vocalisation of either סֵף or assuming that an א aleph has fallen away due to haplography and it should therefore be understood as ף ס the ,אֶאְֶ verb סֵף can still be understood as a Qal imperfective verb in the first person common singular א that can be translated as I will gather. With the combination of the Qal infinitive absolute the tautological infinitive would give rise to the translation I will surely gather or I will gather up/ away. Olivier (1973:5) presents extensive research on various possible explanations for the combination of a Qal infinitive absolute and a Hiphil imperfective first person common singular verb. His commentary includes criticising Verhoef, his promotor for the M.Th. dissertation. Verhoef apparently wanted to translate this expression as I will sweep away entirely, but Olivier (1973:5) argued against it. Olivier (1973:5) suggested a translation that uses the simple future tense in Afrikaans to indicate the finality of the judgement being announced. He suggests that the Qal infinitive absolute ף ”has the sense of “once again אָס whilst אָסֵף as a Hiphil Imperfective may be retained. Olivier (1973:3) suggested a translation that would read I will once again sweep away everything from the face of the earth, says the 27 Lord. The first translation draft during the first phase of the official translation process read Ek gaan alles van die aarde af wegvee, sê die Here (“I will sweep away all from the earth, says the Lord”). As of the second phase of translation drafts, the reading was changed to the reading as it was published in 1983. Ultimately AFR83 translates this expression as a future tense Ek gaan ’n einde maak aan alles op aarde (“I will make an end to all things on earth”). AFR83 seems to render only one of the two verbs in translation. AFR83 contains no sense of two verbs present in the source text, no matter from which roots they are assumed to be derived, working in combination. It is interesting that AFR33 conveyed the presence of two different verbs, although it is not clear how the translators of AFR33 understood the roots or the combination of the grammatical forms. AFR33 translates this expression as Saamskraap, wegraap (“scrape together, reap away”) without a conjunction, the same as the Hebrew text. AFR83 seems to render only one verb. If so, the translators have used the strategy of omission (Naudé 1999:79). Whether the translators felt that their translation does indeed convey the meaning of the combination of the emended reading (i.e. סֵף ף א it not clear. What is clear, is that AFR83 creates the impression (אָס of conveying only one verb and not two as is written in the ST and the text-critical apparatus does not suggest that one of the two verbs should be deleted. This is an example where the translators deviated from both the Hebrew text of BHS and its text-critical apparatus. By analogy, the translators must have used a different source apart from BHS to establish their translation. This demonstrates the presence of a complex source, a complex incipient sign system, that would have consisted of at least more than one text apart from the stated ST. It is not clear in what way or if the NEB and GNB would have influenced the translators of AFR83 directly in this case, but the similarity between AFR83 Zephaniah, NEB, and GNB of this text is remarkable because they all render only one verb in translation. 3.3. Zephaniah 1:4 ב ָ֔ ר ה זֶה֙ אֶת־שְאֵָ֣ ום ה מָק ִ֤ ן־ה י מִּ תִִּּ֞ כְר ִּם וְהִּ י יְרוּשָלָ ושְבֵֵ֣ ל כָל־י ָ֖ ה וְע ל־יְהוּדָָ֔ י֙ ע י יָדִּ יתִּ ִ֤ ם וְנָטִּ ל אֶת־שֵֵ֥ ע ים׃ ָֽ הֲנִּ כ ם־ה ים עִּ ָ֖ כְמָרִּ ה AFR83: Ek gaan my hand uitsteek teen Juda en teen al die inwoners van Jerusalem, Ek gaan uit hierdie plek uitroei wat daar nog van Baäl oor is: die naam van die afgodspriesters, van alle priesters: [I will stretch out my hand against Judah and against all the inhabitants of Jerusalem, I will exterminate from this place all that still remains of Baal: the name of the idol priests, of all priests.] 28 KJV: I will also stretch out mine hand upon Judah, and upon all the inhabitants of Jerusalem; and I will cut off the remnant of Baal from this place, and the name of the Chemarims with the priests; SV: En Ik zal Mijn hand uitstrekken tegen Juda, en tegen alle inwoners van Jeruzalem; en Ik zal uit deze plaats uitroeien het overblijfsel van Baal, en den naam der Chemarim met de priesters; AFR33: En Ek sal my hand uitstrek teen Juda en teen al die inwoners van Jerusalem; Ek sal ook uit hierdie plek uitroei wat van Baäl oorgebly het, die naam van al die afgodspriesters onder die priesters; NBG-51: Ik zal mijn hand uitstrekken tegen Juda en tegen alle inwoners van Jeruzalem, en Ik zal uit deze plaats uitroeien het overblijfsel van Baäl en de naam der afgodsdienaren met de priesters, AFR53: En Ek sal my hand uitstrek teen Juda en teen al die inwoners van Jerusalem; Ek sal ook uit hierdie plek uitroei wat van Baäl oorgebly het, die naam van al die afgodsdienaars saam met die priesters; NEB: I will stretch my hand over Judah and all who live in Jerusalem. I will wipe out from this place the last remnant of Baal and the very name of the heathen priests. GNB: I will punish the people of Jerusalem and of all Judah. I will destroy the last trace of the worship of Baal there, and no one will even remember the pagan priests who serve him. AFR83 translates ים ָ֖ כְמָרִּ ם ה י as the object of the verb אֶת־שֵֵ֥ תִִּּ֞ כְר ים but renders וְהִּ ָֽ הֲנִּ כ ם־ה as if it עִּ were part of the construct chain, omitting the preposition עם and assuming that הכהנים also links to שם. This has the sense that I will cut off the name of the idolatrous priests, the name of all the priests. There are several text-critical considerations in this verse. At least two of these seem to have caused some differences between translations because the text-critical problems result that the meaning of the ST is not clear (Nord 2005:168). According to the text-critical apparatus of BHS, which gives the same information as the apparatus of BHK3 in this case, the phrase עם־ is omitted by the Septuagint. The other problem is that the meaning (”with the priests“) הכהנים of הכמרים is unclear (Ben-Zvi 1991:67). KJV and SV transliterate this word as Chemerim, whilst the other translations vary between interpreting it either as the worshippers of idols or as the priests of idols. AFR83 omits the waw by translating ונטיתי only as Ek gaan uitsteek (“I will stretch out”) without the conjunction. This does not necessarily mean that AFR83 explicitly followed a different Hebrew reading than its ST in this case. Since the theory of functional equivalence 29 is more concerned with the transfer of meaning than the forms of the ST, it is possible that the translators of AFR83 did not believe that the omission of the conjunction would diminish the meaning for the target audience and could justify its omission in that way. The translation of הכמרים seems to have presented problems for the various translations that might have been incipient to AFR83 Zephaniah. The translation of הכמרים is significant because it hints at some of the sources that the translators of AFR83 Zephaniah could have used to establish its meaning. Between AFR33 and its revision (AFR53) there seems to have been a shift in interpretation. AFR33 translates הכמרים as the priests of idols, whilst AFR53 translates it as the servants or worshippers of idols. Due to the proximity of the dates of publication of NBG-51 and AFR53, this might have been because of the influence of NBG-51 upon AFR53. NBG-51 followed the interpretation that הכמרים refers to the worshippers or servants of idols. Olivier’s (1973:14) explanation of this word is quite significant since his research was directly accessible to the translators of AFR83 Zephaniah. Olivier (1973:14) does not seem to entertain the possibility of other translations for הכמרים in detail. He seems to believe that the meaning of הכמרים is not obscure at all even though it appears only four times in the Hebrew Bible. According to Olivier (1973:14) every appearance of this word in the Hebrew Bible (Zephaniah 2:4; 2 Kings 23:5; Hosea 4:4; Hosea 10:5) refers to the priests of idols and that is why this may be accepted as the translation equivalent. AFR83 Zephaniah does represent a deviation from BHS in their rendition of יםכהנ עם־ה . Some possible incipient texts such as KJV, SV, and AFR53 render it literally as with the priests with the sense that the priests (הכהנים) will be annihilated along with the group interpreted either as idol priests or idol worshippers (הכמרים). NEB omits עם־הכהנים following the Septuagint (Brockington 1973:262) as does the GNB. AFR33 translates it as onder die priesters (“among the priests”) with the sense that הכמרים is a sub-unit that resides among הכהנים who will be annihilated. The sense of AFR33 rendition can be translated as that הכמרים will be annihilated from among the priests. Olivier (1973:14) refers to this interpretation and makes the judgement that such a translation will create problems for the perceived rhythmic meter of the verse, but he does not elaborate further. Be that as it may, some of the above translations presuppose that they included the preposition עם (“with”) in their reading of the text, although they rendered it in different ways. 30 AFR83 Zephaniah’s translation presents a different interpretation of the syntactical relationship between הכמרים and יםהנ הכ than that contained in the Hebrew text. There is more than one possibility of interpreting how AFR83 read their ST in this case. If AFR83 of Zephaniah 1:4 understood עם־הכהנים as something similar to I shall annihilate the name of the idol priests (הכמרים), in other words/ that is, all the priests (הכהנים), then AFR83 conveyed the prepositional phrase עם־הכהנים (“with the priests”) as an appositive phrase, omitting the preposition עם, thus changing the grammar between the ST and the translation. If AFR83 of Zeph. 1:4 is understood as something similar to I shall annihilate the name of the idol priests, the name of all the priests, then they read הכהנים as if it were part of the construct chain that originates from the previous phrase, omitting the preposition עם. Whichever might represent the understanding of the translators of AFR83 Zephaniah, they conveyed the prepositional phrase עם־הכהנים by using a different grammatical construction in the target text. The translators of AFR83 have followed the translation strategy of transposition which deviates from the Hebrew text by using a different grammatical structure than that of the source text. The translators did not follow the suggestion in the text-critical apparatus of BHS that עם־הכהנים should probably be omitted based on the Septuagint which represents an older (but not necessarily better) reading. Another remarkable feature of Zeph. 1:4 in AFR83 is the addition of a word that is not present in the ST. AFR83 reads Zeph. 1:4 as all the priests. Seen in this way, the translators of AFR83 have employed the translation strategy of addition (Naudé 1999:79). Whichever of these readings represent the actual intention of the translators, is difficult to ascertain. What is clear, is that the translation deviates from its ST both in the way that it rendered the Hebrew text and in that they did not follow the text-critical apparatus in this case. This represents an example where the translators must have drawn from some other source than ST. 3.4. Zephaniah 1:7 יו׃ יש קְרֻאָָֽ ֵ֥ קְדִּ ח הִּ ין יְהוַָ֛ה זֶָ֖ב ִ֧ י־הֵכִּ ָֽ ה כִּ ום יְהוָָ֔ וב֙ י ֵ֣ י קָר ִ֤ ה כִּ נֵָ֣י יְהוִּ פְנֵָ֖י אֲד ס מִּ ה ַ֕ AFR83: Wag in stilte op die Here God! Die dag van die Here is naby! Hy het 'n maaltyd voorberei, en dié wat Hy genooi het, gewy. [Wait upon the Lord God in silence! The day of the Lord is near! He has prepared a meal, and those whom he has invited, he has consecrated.] 31 KJV: Hold thy peace at the presence of the Lord GOD: for the day of the LORD is at hand: for the LORD hath prepared a sacrifice, he hath bid his guests. SV: Zwijgt voor het aangezicht des Heeren HEEREN; want de dag des HEEREN is nabij; want de HEERE heeft een slachtoffer bereid, Hij heeft Zijn genoden geheiligd. AFR33: Swyg voor die Here HERE! Want die dag van die HERE is naby; want die HERE maak ’n offer klaar, Hy heilig sy gaste. NBG-51: Zwijg voor het aangezicht van de Here HERE, want nabij is de dag des HEREN; want de HERE heeft een offermaal bereid; Hij heeft zijn genodigden geheiligd. AFR53: Swyg voor die Here HERE! Want die dag van die HERE is naby; want die HERE het 'n offer berei, sy gaste geheilig. NEB: Silence before the Lord GOD! For the day of the LORD is near. The Lord has prepared a sacrifice and has hallowed his guests. GNB: The day is near when the Lord will sit in judgment; so be silent in his presence. The Lord is preparing to sacrifice his people and has invited enemies to plunder Judah. AFR83 translates הס as an imperative wag (“wait!”). The analysis of this verb is not clear. In Hebrew it is not clear whether the word is an interjection (Olivier 1973:26) with the purpose of onomatopoeia in mind (i.e. similar to hush! in modern English as an interjection), whether it is a Piel imperative based on the root הסה or whether it is derived from the root הסס, understood as an imperative in the Qal form. Be that as it may, according to Ben-Zvi (1991:79), the meaning is ultimately inferred from the context. Of the possible incipient texts, it is only NEB that translates הס as an interjection as silence followed by an exclamation mark. AFR33 seems to have imitated the Dutch SV with the word swyg (“be silent!”), as an imperative, which has been retained in both the AFR53 revision and the NBG-51. The singular word swyg in Afrikaans and swijg in Dutch might have the characteristic of an interjection despite being imperative verbs, but the KJV and GNB translate הס as a verbal phrase as hold thy peace and be silent respectively. The translators of AFR83 Zephaniah render הס as wag in stilte (“wait in silence”) as an imperative verb followed by an adverbial phrase. The translators of AFR83 Zephaniah have possibly been guided either by their adopted translation approach or by the varying translations that might have influenced them. The question remains whether this translation represents an instance where the translators of AFR83 Zephaniah deviated from the source text or not. In the meeting of the Translation Commission of 2nd February 1971, one of the guidelines established by the Commission is that 32 which is implicit in the source text must be made explicit in the target text. הס could be interpreted as either an interjection or an imperative verb. The translators decided to render it explicitly as an imperative verb. However, the translators have also employed the strategy of addition (Naudé 1999:79) in this case. The word הס does not necessarily mean to “wait”. Be it understood as an imperative from different roots or as an interjection, the meaning is clearly to invoke silence, not the act of waiting. Yet the translators have rendered it so. The translators have decided to use a different grammatical structure to convey the meaning of this word in the ST. The translators have translated the single word הס in ST as an imperative verbal phrase and with a different meaning in the TT from the original word in the ST. AFR83 retains the meaning of הס (“silence”) by means of an adverbial phrase in the target language. The imperative is translated as wag (“wait!”), which is not reflected in the source text, followed by an adverbial prepositional phrase in stilte (“in silence”), modifying the Afrikaans verb wag (“wait”). The translators of AFR83 Zephaniah seem to have used a combination of the strategy of addition and transposition in this case (Naudé 1999:79). Although it is not certain what informed the translators’ decision, it is clear that the translation represents neither the Hebrew text, nor the suggestions of the text-critical apparatus of BHS. They have deviated from both the Hebrew text and the text-critical notes of BHS in this case not only by adding something in the translation that is not present in the source text, but also by changing the grammatical structure from the source text into the target text. Incipient to this translation is perhaps not another text as such, but simply the desire of the translators to produce a readily understandable translation. This in an illustration of what Marais (2019:45) calls an incipient sign system that includes not only texts, but other meaning-bearing elements as well. In this case not necessarily another text, but the translators’ subjective desire for a comprehensible translation could have been the main incipient factor. Whatever might have informed the decision of the translators of AFR83 Zephaniah in this case, it is clear that they deviated from their ST. This illustrates that the translators would have drawn from a source other than just their explicitly stated source text and that the argument for the presence of a complex incipient sign system is justified in this case. 3.5. Zephaniah 1:9 ג ולֵַ֛ ד ל כָל־ה ִ֧ י ע דְתִֹּּ֗ ה׃ וּפָק רְמָָֽ ס וּמִּ ם חָמֵָ֥ נֵיהֶָ֖ ית אֲד ים בֵֵ֥ ַ֛ לְאִּ מְמ ָֽ וּא ה ה ום ה י ֵ֣ ן ב פְתָָ֖ מִּ ל־ה ע AFR83: Daardie dag sal Ek almal straf wat uit bygeloof nie eens op 'n drumpel trap nie en wat die huis van hulle Koning met geweld en bedrog vul. 33 [On that day I will punish all who, out of superstition, do not even tread upon the threshold and who fill the house of their King with violence and corruption.] KJV: In the same day also will I punish all those that leap on the threshold, which fill their masters' houses with violence and deceit. SV: Ook zal Ik ten zelven dage bezoeking doen over al wie over den dorpel springt; die het huis hunner heren vullen met geweld en bedrog. AFR33: Ek sal ook op dié dag besoeking doen oor almal wat oor die drumpel spring, wat die huis van hulle here vul met geweld en bedrog. NBG-51: Ook zal Ik te dien dage bezoeking doen over allen die over de drempel springen, die het huis hunner heren vullen met geweld en bedrog. NEB: I will punish all who dance on the temple terrace, who fill their master’s house with crimes of violence and fraud. GNB: I will punish all who worship like pagans and who steal and kill in order to fill their master’s house with loot. The KJV represents the literal translation of this verse. The possible differences in the translations originate from the various ways to translate the preposition על which KJV translates as on, with the sense that those against whom judgement is proclaimed in Zeph. 1:7 do indeed step upon the threshold. NEB follows this interpretation as well. The Hebrew word על can be translated in more than one way (e.g. on, upon, over, above). This is one of the things that could possibly have given rise to the very different interpretations of this verse when the various other possible incipient texts are compared with each other. The Dutch SV translates this word as over. In Dutch this word indeed means “over” with the sense that the person against whom judgement is proclaimed in Zeph. 1:7 steps “over” the threshold (i.e. he does not tread upon the threshold, but steps over it, he avoids stepping upon it). Both AFR33 (AFR53 corresponds verbatim with AFR33 in this case) and NBG-51 follow this interpretation as well. AFR83 follows the same interpretation as SV, AFR33, and NBG-51, but the translators decided to convey the positive statement of the Hebrew (“they leap over the threshold”) as a negative statement in translation that “they do not trample upon the threshold.” This is an example of the strategy of transposition and addition by the translators of AFR83 Zephaniah (Naudé 1999:79). There is no negative particle in the Hebrew present in the source text and the Hebrew conveys a positive statement. AFR83 does the opposite (i.e. the translators insert the negative 34 particle in Afrikaans and translate it as a negative statement), although the meaning of both the positive statement in the Hebrew and the negative statement in Afrikaans is the same. Another example of the strategy of addition, is that the translators of AFR83 Zephaniah added the words uit bygeloof (“out of superstition”) which is not present in the source text at all. The sense of the translation is that the Lord will punish those who do not step upon the threshold because they are superstitious. This could originate from the translation policy that the translation should make explicit that which is implicit in the text by adding, by incorporating the reason for the action into the text of the translation itself. In their rendition of בית אדניהם (“the house of their lord”) the translators followed the strategy of specification by translating it as die huis van hulle Koning (“the house of their King”). The word אדון (“lord”) can certainly be understood as a king, but not necessarily. It would have to be inferred from the context. The fact that AFR83 Zephaniah 1:9 capitalises the word for “king” indicates that the translators wanted to guide the reader that those who are going to be punished are those who fill the house of the Lord, the temple of the Lord, with violence. This is not necessarily the meaning. According to Olivier (1973:36) it could refer to the temples of idols that are being filled with violence and that the act of not stepping over the threshold refers to some superstition within that context. The readings adopted by the GNB, NEB, the Dutch and other Afrikaans translations seem to leave open the interpretation about who exactly the lords and their houses are that are being referred to in Zephaniah 1:9. In Zephaniah 1:9, the translators of AFR83 Zephaniah have employed the strategies of addition, transposition, and specification (Naudé 1999:79). The strategy of addition indicates that the translators deviated from the Hebrew text of BHS as well as its text-critical apparatus. They also deviated by translating a positive statement in the Hebrew text as a negative statement in translation. This is an example of where the translators must have drawn their translation from a source other than BHS. This can serve as a demonstration that the source of AFR83 translation of Zephaniah 1:9 was not exclusively BHS, but that it points to a complex source that would have consisted of at least more than one text. 35 3.6. Zephaniah 1:12 מְרֵיהֶָ֔ ל־שִּ ים֙ ע פְאִּ ק ָֽ ים ה ל־הָאֲנָשִֹּּ֗ י ע ֵ֣ דְתִּ ות וּפָק נֵר ִּם ב ָ֖ שׂ אֶת־יְרוּשָל פֵֵ֥ יא אֲח הִָּ֔ ת ה ים֙ וְהָיָה֙ בָעֵֵ֣ מְרִּ ם הָא ָֽ ם לְבָבָָ֔ ׃בִּ ע א יָרֵָֽ יב יְהוָָ֖ה וְל ֵ֥ ֵ֥ א־יֵיטִּ ל ָֽ AFR83: Daardie tyd sal Ek Jerusalem met lampe deursoek en die mense straf wat in hulle selftevredenheid dink: die Here doen nie goed óf kwaad nie. [At that time, I will search Jerusalem with lamps and punish those who think in their indifference/ self-confidence/ self-contentment: the Lord does neither good nor bad.] KJV: And it shall come to pass at that time, that I will s