On 29 November 2020, the new Afrikaanse translation, Die Bybel, 2020-Vertaling, was launched in Bloemfontein (hereafter referred to as the 2020 Afrikaanse Vertaling). This marked the publication of other complete Bibles in Afrikaans: the 1933/1953 Die Bybel in Afrikaans; the 1982 Lewende Bybel; the 1983 Die Bybel in Afrikaans; Die Boodskap of 2002; the 2007 Nuwe Lewende Bybel; and the 2007/2008 Die Bybel vir Almal. This is not a complete list.¹ With the significant growth of Bible translation theory and practice over the past 60 to 70 years, new theories led to new approaches and new styles of translation, initially accompanied with an assertion that the new method is superior, and the new translation could and should replace previous publications. That in turn led to a reverse situation where ‘literal only’ or ‘literal-superiority’ claims were made (see e.g. Grudem et al. [2005]; Fee & Straus 2007; and Brunn’s helpful discussion [2013] in this regard). Over time, as Bible-translation theory became more sophisticated and translation studies in general expanded, more nuanced perspectives prevailed that different styles or different types of translation might meet different requirements of Bible users. Hence, the rise of niche² translations – translations that fulfil a certain purpose and function in contrast to other purposes and functions.

The purpose of this article is to point out some translation type categories on a continuum, and that the different Afrikaans Bibles could indeed fit into a continuum of translation types. In addition, the different versions in Afrikaans are not only points of a type of continuum, but may also point to end-user niches for each in the Bible market.

The article develops in the following way: firstly, an overview is given of how translation studies developed to the point that several translation types can be discerned. This is followed by defining four instead of only two translation types with criteria that differentiate each one. Comparing Afrikaans versions according to these criteria is a logical next step. The final argument is to link the four types with four probably niches for end-user scripture engagement, and some conclusions based on a niche analysis.

Translation type, translation model and translation style are all terms that point to the same object of study in translation studies. Translation type will consistently be used in this article. A type is a classification based on categories and common criteria; or ‘a group of things that have particular features in common’ (Collins Cobuild English Dictionary 1995). In the categorisation of types, groupings of Bible translations that share common features and characteristics will be the object of this study.

Before the introduction of the dynamic-equivalent translation theory of Eugene Nida (1964), and Nida and Taber (1969), most translations followed the formal equivalence model – also called literal translation. In the English-speaking world, the King James Version (KJV [1611]), the Revised Standard Version (RSV [1952]), and New American Standard Bible (NASB [1971]) are literal translations. In Dutch, it was the Statenvertaling (SV) of 1637, and the Nieuwe Vertaling of 1951, in Portuguese the version of João Ferreira de Almeida (NT in 1681, Bible in 1821) and its many revisions (1993; 1995; 2017). In Afrikaans, the 1933 edition with its 1953 revision also falls in this category.

This was changed by the work of E.A. Nida and his fellow scholars with the introduction of a second translation type called dynamic-equivalent translation. A distinction was made between form and meaning, and this distinction was given academic authenticity. Beekman and Callow (1974:25–35), as well as Mildred Larson (1984:3–6) essentially followed the same line, except that dynamic-equivalent translations are called idiomatic translations.

The Bíblia Sagrada Boa Nova (1993) and Bíblia Para Todos (1993) versions in Portuguese, the Groot Nieuws voor Uw translation in Dutch (1982/1996), and the 1983 translation in Afrikaans (referred to as the 1983 Nuwe Vertaling in this article) are all versions of this type. In English, the publication of the Good News Bible (GNB) or the Today’s English Version in 1966 opened the way for this type to be dominant in the Bible translation world until the 1990s with the dichotomy of the two types already mentioned: literal versus dynamic-equivalent or idiomatic remains popular in translation studies to this day.

The theory was further fine-tuned with the emergence of the functional-equivalence theory, building on and improving earlier theories (De Waard & Nida 1986). However, from the 1990s onwards, dynamic-equivalent translation was challenged by new linguistic developments, especially in pragmatics and with the pragmatics theory of Relevance Theory (the application of Relevance Theory to Bible translation of Sperber & Wilson [1986]; and Ernst-August Gutt [2000]). This led to a third type. As described and applied to Bible translation by Gutt (2000), communication happens through inference, and translation falls into two general types: direct translation and indirect translation. Analogous to direct and indirect speech, direct translation is ‘retelling’ the original as close as possible, whereas indirect translation is more like describing the meaning. As Van der Merwe (2016:434) pointed out, direct translation is a new ‘type’ of translation, without using the term, but giving some criteria. Direct translation is different from the traditional literal or formal correspondence translation, but it is not the same as functional-equivalent either. It lies somewhere in-between.

Much has already been written concerning the application of Relevance theory on Bible translation in the context of Afrikaans,³ especially during the process of researching a new type of translation. However, Relevance Theory is not the only influence in translation studies moving away from dynamic/functional- equivalence and the rise of a third type. Discourse studies developments in information structure (topic and focus), sociolinguistics and cognitive semantics influenced the way translators deal with word order, honorifics, social context and lexical choices.⁴ For instance, lexical choices were either concordant (literal translations) or context-determined (dynamic translations). Another influence on translation types has been Skopos theory and, especially, the work of Christiane Nord (2018) on functional translation. The quality of a translation is not determined by some universal standards applicable to all cases, but rather by the skopos [purpose] or function intended. Skopos theory (Nord 2018) has shed light on translations developed for certain specified audiences, be it young readers, the deaf, readers in a mission situation, et cetera. Arguing for only one type as the correct one is increasingly disregarded as an unacceptable, prescriptive position.

One can illustrate this development of Bible translation types as a zigzag pendulum, swinging from formal correspondent to dynamic-equivalent, and from dynamic-equivalent to idiomatic literal⁵ (using a Brunn [2013] term). The very free swing has been inspired by both the dynamic-equivalent and the idiomatic literal type.

The chronology of publications, as mentioned above, confirms the trend: first the formal correspondence versions, then the dynamic/functional-equivalent versions and then the more idiomatic literal or direct translations. Highly dynamic and free translations were around from the beginning of dynamic-equivalent translations as a more extreme, freer and more explicit translation type. Figure 1 lays out graphically the historical developments of the different types.

Beekman and Callow (1974:19–32) were the first to propose four translation types: (1) unduly literal; (2) idiomatic literal; (3) idiomatic; and (4) unduly free.

They evaluated types 1 and 4 as unacceptable translation types. There are also other typologies, most notable in the functionalist school in translation studies. Newmark (1988:45) distinguishes the two main groupings: source language (SL) emphasis; and target language (TL) emphasis, and then proposed several translation types under each main grouping:

In total, Newmark (1988) distinguishes eight types. The two-fold distinction, SL emphasis and TL emphasis, agree with most theorists in translation studies. The SL emphasis categories, except the interlinear, word-for-word, seem to be on a spectrum from very formal correspondent to less so with more focus on meaning and context whilst still having an SL emphasis. The same is true for the TL emphasis categories. It is not clear, even from Newmark’s definitions, how faithful and semantic SL emphasis translations differ from each other. It seems to be that it is only a matter of degree. The same can be said of idiomatic and communicative translation. The difference between ‘semantic translation’ and ‘communicative translation’ is not clear either. They seem to be very similar – again only a matter of degree. The labels of the subcategories raise a more serious question, however. They overlap in meaning and several can be used on either side of the two main groupings. The labels of faithful translation, semantic translation, idiomatic translation and communicative can be attached to both columns, for instance. But it must be said that Newmark was breaking new ground when he, as one of the first translation scholars, went beyond the two main type categories and even beyond the four types of Beekman and Callow (1974).

The Bible Society of South Africa, whose publications in Afrikaans will be compared in this article, distinguishes between formal translation, translation in contemporary language, and easy-to-read versions. However, some of these categories overlap. The 2020 translation is not only formal, but in contemporary language, and the easy-to-read version, die Bybel vir Almal, is also in contemporary language. Like Newmark’s categories, the categories of classification overlap. Hence, the need for more criteria to distinguish the types.

In my 2007 article, ‘The four translation types and criteria to distinguish them’, I raised the question that there is a need to distinguish more than only two general translation types: literal versus idiomatic or functional-equivalent. I shall briefly summarise my proposed categorisation of the four types.

First, the question must be asked, why four types instead of only the original two (literal vs. dynamic or idiomatic [Nida & Taber 1969; Beekman & Callow 1974:19-32]; documentation versus instrumentation [Nord 2018], etc.), as still maintained in general translation studies. There is widespread consensus that the two main types remain a valid distinction. The above discussion has provided a brief overview of the rise of four different translation types. The inadequacy of only two types will be developed further in this section:

As translation types are on a continuum, differing from each other by degree, it is necessary to identify criteria to sub-differentiate and rename the types. I still distinguish two main types, but to capture this more elastic typology, reflecting the degrees of literalness or idiomaticity, the one main type is called more literal translation, and the other more idiomatic translation:

For each main type, two subtypes are suggested and named. The following four verbs are suggested for the four types:⁶ corresponding; resembling; clarifying; and simplifying.

A more literal translation is more source-language oriented, and there are two subtypes:

A more dynamic/idiomatic translation is more receptor-language oriented, and there are two subtypes:

Why all the new terms? The terms need to be accessible for the Bible public and Bible users to quickly grasp what some of the basic issues are in terms of translation type choices.

It is now necessary to differentiate the four types. Eight criteria are proposed, based on syntactic (1–3), semantic (4–6) and discourse-pragmatic features (7–8):

Each of the four types will now briefly be characterised.

Below follows an overview of Bible translations in English and Dutch, comparing versions⁷ according to the translation type continuum as stated above (see Table 7).

The history of the Bible and Bible translation in Afrikaans have been described elsewhere (Van der Merwe 2012; 2016; 2020; Naudé & Van der Merwe 2002; Naudé 2005). The purpose of this comparison of five versions in Afrikaans is to show how they differ from each other in light of the typology suggested above, and that, in Afrikaans, we now have a version in each of the four types.

The 1933/1953 version is a type 1 translation, and the 1983 version a type 3. It is also not too difficult to show that Die Bybel vir Almal meets the description of a type 4 simplifying translation. But where on the continuum is the new 2020 Afrikaanse Vertaling? A full type 1, or maybe more to the right on the continuum towards type 2? Or maybe in between type 1 and type 2 with characteristics of each? Or maybe a full type 2, leaning a little towards type 1? The SL orientation and the inclusion of the original context and its cognitive environment make it a more literal translation, but, at the same time, the contemporary and clear language points to a sharp inclination to type 2 resembling translation. According to the criteria of each type as laid out above, the 2020 Afrikaanse Vertaling leans towards type 2, but with some type 1 characteristics.

For the 2020 Afrikaanse Vertaling the definition of a type 2 translation is therefore proposed. The skopos of the translation can be described as in-between type 1 and type 3. As a direct translation, it attempted to overcome the distance of time and context in a few cases, keeping the formal correspondent rendering (e.g. the idiom julle lendene omgord in 1 Petrus 1:13. For some difficult words, military ranks and measures the original was transliterated and a footnote provided to explain the meaning. An example is parakleet in John 14:16. In most cases, a series of adjustments were made in line with the eight criteria listed above. Instead of full one-for-one concordance of lexical items, partial concordance has been implemented. Sentences have been shortened where better reading demanded it, original transitions and connectives, particles such as the Hebrew hinneh [look, behold], as well as marked word-orders were not rendered in a full concordant, one-for-one way, but were translated with a sensitivity for the meaning and function of the original. The 2020 translation is more in line with the type 2 criteria than with the type 1 criteria, but, of course, type 1 and 3 renderings also feature.

The 2020 Afrikaanse Vertaling came to fill a gap in Afrikaans, a good, standard type 2 translation prepared according to clear skopos requirements and based on the latest linguistics and translation studies.⁸

Five Afrikaans versions are compared according to this classification of types⁹ (Table 8).

Here follows specific comparative examples of the four types in Afrikaans.

‘Highly literal’ and ‘unduly free’ are concepts used in Beekman and Callow (1974:21–23) to refer to the two fringe types. They define these translations as either too literal to be understood well or too free and at risk of not being accurate. ‘Unduly free’ and ‘highly literal’ are value statements and therefore can be interpreted as prescriptive. There are several points in this regard to be made:

In light of this preamble, a concluding niche analysis may be helpful.

The conclusion of this analysis is that the translations in one language may be on different points of the continuum. They are not necessarily in competition with each other, but rather to be seen as complementary. The different types are not exclusive. Different end-user requirements and preferences have made it possible for different types meeting specific niche markets.

This, in turn, implies that there will still be a need for type 1 very literal, formal correspondent translations, as well as type 3 dynamic or functional-equivalent clarifying translations. The newer types 2 and 4 have not rendered the older types redundant. The demand for them continues as their availability in bookstores and their sales show, and therefore also the need to continue revising them for new generations. This also means that revisions of type 1 and type 3 translations cannot be excluded.

Could it be that eventually there will be demand for a revision of the 1953 type 1 translation in Afrikaans? There was a demand in the Netherlands for a revision of both the SV and the 1951 Nieuwe Vertaling. Such a revision of the 1933/1953 version will probably only entail a modernisation of the language use for words such as goedertierenheid [loyal love], and maybe to revise a few errors and thus not the full-scale move to a new translation theory as was the case for the 2020.

The same can be said for the 1983 Nuwe Afrikaanse Vertaling – a type 3. In many of the European languages, those type 3 versions were also revised or parallel versions published. An example is the NLT that appeared next to the GNB – different, but typologically close.

A second practical implication of this type and niche analysis is that multiple Bible use, in the sense of a person or a group using different types and not one type exclusively, can and even should be encouraged for the wider benefit of the Body of Christ and its mission in the world. Bible users will benefit from reading or listening to the Scriptures of complementary types.