Double or half reading, double or full meaning: Amphibological and anacoluthic syntax through the lens of Qur’an translators

Full Text

1. Introduction

The idea of this study grew out of a first-hand English translation of one of the oldest but most reliable sources on the Muslim conquests of Syria (Hassanein & Scheiner 2020). Therein the translators adopted an SL¹-oriented philological approach to the (un)translatability of al-iltifa:t (‘deictic shift’) and fuqda:n at-tata:bu^c (‘anacoluthon’) as proper manifestations of ambiguous and anacoluthic Arabic syntax believed to be modelled basically on Qur’anic syntax (see Kha:tir 2000: 42, Kiss & Alexiadou 2015: 100). The present study limits its scope to two under-examined syntactic phenomena in Qur’anic Translation Studies (QTS). One is al-labs an-nahwi: (‘syntactic ambiguity/amphibology’), which figures prominently in a:ya:t at-ta^ca:nuq (‘inter-embracing verses’). These Qur’anic verses are superscripted by two inter-embracing pause signs (^{∴ ∴}), whereby reading and meaning differ based on which pause sign the Qur’an reciters pause at (see Al-Harbi: 2004). The other is fuqda:n at-tata:bu^c (‘anacoluthon’) which is defined in Arabic rhetoric as an abrupt shift to a second sentence before the first is meaningfully completed (see Ha:mid & Qandi:l 2019: 100). These two syntactic phenomena typify al-ija:z bi-l-hadhf (‘brevity by ellipsis’), which gruellingly challenges the Qur’an translator.

The Qur’anic text attracts so riveting and scrupulous research that each of its linguistic phenomena may be subjected to many scholarly endeavours (see Mohaghegh & Pirnajmuddin 2013, Abdul-Ghafour et al. 2019, Raoufkazemi et al. 2020, Alduhaim 2021). Abdul-Raof (2001: 68) argues that the Qur’anic language is characterised by rhetorically, syntactically, semantically, phonetically, and pragmatically idiosyncratic and prototypical features which render it roughly (un)translatable.² Therefore, advocates of the untranslatability of the Qur’an contend that however professionally gifted the translators are, there is a slim opportunity that they are able to transfer the dynamic effect of utterances as they are in the Qur’an (El-Hadary 2008: 39). On the translation of the Qur’an, Naudé (2010: 289) expounds that no existing translation in English reflects the language-dependent nature of the performance of the Qur’an or mirrors its majesty and aesthetic appeal. He (2010: 286) states that what is required is a target-oriented strategy to serve a new skopos independent of that of the original rather than strive for equivalence.

A constellation of translation scholars raise counterarguments that prioritise the ST over the TT. Newmark (1988: 220) advises the translators to reproduce the intentional ambiguity whenever possible, and if its reproduction is impossible, they may translate the most probable sense and footnote the less probable if they believe it to be important. Baker (1992) gives equivalence prominent significance in translation to the extent that she tackles problems arising from a lack of equivalence at different linguistic levels. Pym (1995: 168) asserts that theorists who oppose equivalence do not have “a restrictive definition of translation”. Nida (2001: 1) regards faithful equivalence in meaning as an additive, not competitive, factor besides clarity of form and elegance of content. Dickens et al. (2016: 16) state that the inability to achieve equivalence is a translation loss. Mustafa (2019: 29) considers equivalence to be so essential a component of translation that its absence presents a problem in giving it a tangible definition.

The present study seeks to examine how Qur’an translators have handled the syntactic complexities stemming from the two understudied issues in focus: amphiboly and anacoluthon. The rationale for laying special emphasis on these two particular phenomena develops from the literature review which reveals that they pass untended in Qur’an Translation Studies and the postulation that they bring Qur’an translators into a translational dilemma: a quandary over explicitation, implicitation, or neutralisation. The specific objectives are

1. to investigate how the Qur’an translators resolved syntactic ambiguities during their rendering of inter-embracing verses signalled by the pause sign (^{∴ ∴}), and
2. to explore how they sequentialised anacolutha while translating anacoluthic verses marked by the elliptical sign (…).

The compendious inquiry is whether or not they disambiguated amphiboly and sequentialised anacoluthon while translating Qur’anic syntax.

2. Preliminaries

This section introduces the target reader to the state-of-the-art literature review and cutting-edge theoretical preliminaries to the two issues under scrutiny.

2.1. Waqf at-ta^ca:nuq (‘inter-embracing pause’) in Arabic

Waqf at-ta^ca:nuq or at-taja:dhub (‘inter-embracing or inter-attracting pause’) creates syntactic ambiguities in a limited number of Qur’anic verses referred to as a:ya:t al-mu^ca:naqa (‘inter-embracing verses’), which are superscripted by a double pause sign (^{∴ ∴}) (see Omer 1997, Al-Harbi: 2004). In these verses, a pause at one sign necessitates a non-pause at the other, leading to different interpretations and translations. The two signs work on a complementary or binary basis in that they are mutually exclusive in Qur’anic tajwi:d (‘articulate recitation’). A pause at either sign requires a continuation at the other. Al-Harbi: (2004: 4) argues that waqf at-ta^ca:nuq is a subjective and problematic choice of reading, interpretation, and translation, claimed to have been first attended by Abu:-l-Fadl Al-Ra:zi: who named it after al-mura:qaba (‘surveillance’) in al-^caru:d (‘prosody’). Figure 1 shows Qur’an recitation signs, of which the interchangeable pause sign comes last on the left.

Figure 1. Signs of al-waqf (‘pause’) in Qur’an recitation
Source. www.google.com by Dna Aryanti

“The issue of identifying ambiguities in the Arabic language has been ignored in almost all the systems that attempted to process Arabic” (Daimi 2001: 346). The case is rather deplorable with respect to Qur’anic syntactic ambiguity that has not received erudite attention in Qur’an translation. Most related to this study is a work undertaken by Al-Ali & Al-Zoubi (2009) on the different meanings triggered by different pausings in syntactically ambiguous (amphibolous) Qur’anic verses. Their purpose was to examine how the Qur’an translators rendered variously meaningful and interpretable verses whose meanings depend upon where the syntactic pause occurs. Findings showed that Qur’an translators opted only for one meaning and ignored the others. In this regard, Newmark (1982: 25) confirms that in all cases of ambiguity, the translators have to consider that the case may be so deliberate that they are obliged to reproduce it in the original or disambiguate it according to the co(n)text, paratextualising, however, the less likely meaning if it might be the intended one. Al-Jarrah, Abu-Dalu & Obiedat (2018) postulate that a good translation of strategic ambiguous structures in CA is not that which interprets the ST, but which leaves the door open for all the interpretations triggered by it, and therefore recommends the direct-translation method because it helps the target reader to draw inferences from given contextual information. In terms of Vermeer’s Skopos Theory, when a text is ambiguous, it can be literally translated but then explained in a footnote (Munday 2008, qtd. in Abdelaal 2019: 3).

2.2. Fuqda:n at-tata:bu^c (‘anacoluthon’) in Arabic

Fuqda:n at-tata:bu^c (‘anacoluthon’), Gr. “lack of sequence or wanting sequence”, is defined in passing in Arabic rhetorical studies as “the sudden change of syntax in a sentence” (Wansbrough 1970: 255), which denotes a break of structure, an incomplete construction, and a disruption of grammar within a sentence. Previous literature on anacoluthon in Arabic is extremely sparse and leaves so wide a gap that I hope to bridge and fill by building on western scholarship and conflating insights thereof into a de facto melting pot (e.g. Kaltenböck 2007, Mieszkowski 2009, Darir 2012, Greene et al. 2012, Greene & Cushman 2016, Lane 2018, Allaithy 2019, Ha:mid & Qandi:l 2019). The most common forms of anacoluthon are the so-called “absolute nominative” and the absence of the second conjunction of a correlative expression known as “particula pendens” when it relates to correlatives (e.g., “both…and”), or as anapodoton (“wanting the apodosis” in Greek) when it relates to the absence of a main clause in a conditional sentence (see Greene & Cushman 2016). A subclass of anapodoton is anantapodoton, in which the subordinate clause is incomplete (see Greene et al. 2012: 46, Greene & Cushman 2016: 11).

Mieszkowski (2009: 648) associates anacoluthon with aposiopesis, in which a sentence breaks off and never continues, and anapodoton, in which a sentence commences with a subordinate clause not followed by a main clause. Anapodoton (Gr. anapodosis “without a main clause”) is said to be the most common form of anacoluthon in the Qur’an (e.g., Justice 1987 qtd. in Ha:mid & Qandi:l 2019: 101, Darir 2012: 10) and hence is subsumed under al-hadhf (‘ellipsis’)–clausal ellipsis in which fi^cl ash-shart (‘the protasis’), the subordinate clause, of a conditional sentence is given whereas jawa:b ash-shart (‘the apodosis’), the main clause, is not mentioned but is inferable from the co(n)text (see Mir 2006: 99, Abdul-Raof 2019: 138, Allaithy 2019: 13).

3. Methodology

Taking no exhaustive stance for space reasons, this section presents a representative sample dataset of Qur’anic verses claimed by a great majority of exegetes, if not by consensus, to be typical of syntactic ambiguity and anapodotic anacoluthon, and a transparent pathway of analysis for the readers to follow. 

3.1. Dataset

Due to space and word limits, the dataset collected and designed for the present study is intended to be representative of the syntactic profiling of the phenomena under scrutiny–a dataset amenable to a rigorous analysis from a contrastive unidirectional translational perspective (SL=L1(A)→TL=L2(E), as put in Klaudy 2005: 13). To achieve the purpose of the study, seven translations that are accessible in the Quranic Arabic Corpus (QAC)³ online have been selected for the contrastive analysis: Sahih International (SI), Marmaduke Pickthall (MP), Yusuf Ali (YA), Habib Shakir (HS), Muhammad Sarwar (MS), Mohsin Khan (MK), and John Arberry (JA). Abdul-Raof (2001: 74) says that an exegesis-free translation which is considered a notoriety for unorthodoxy and heterodoxy requires a consultation of Qur’anic exegeses to decide on the accurate interpretation for transference into the TT. Two linguistically oriented exegeses are employed as counter-reference points for the contrastive analysis: al-Kashsha:f (Az-Zamakhshari: 1998) and at-Tahri:r wa-t-Tanwi:r (Ibn ʿA:shu:r 1984).

3.2. Approach

This study combines the literal and contrastive methods of analysis. The intralingual syntagmatic analysis of clause structure requires a glossing for tree diagramming, bracketing, disambiguation, and sequentialisation. The interlingual syntagmatic analysis of clause structure necessitates using the contrastive linguistic method to compare the structural differences between a pair of languages. This combinative method analyses the (a)symmetry of interlingual transfer, using the unidirectionally contrastive approach (L1→L2, as Klaudy (2005: 15) notates it). On contrastive analysis, Crystal (2008: 112) writes: “A general approach to the investigation of language … particularly … in certain areas of applied linguistics, such as … translation. In a contrastive analysis of two languages, the points of structural difference are identified and … studied as areas of potential difficulty.”

4. Analysis

This section introduces a qualitative contrastive analysis of the dataset of a:ya:t at-ta^ca:nuq (‘inter-embracing verses’) and a:ya:t fuqda:n at-tata:bu^c (‘anacoluthic verses’) as collected from the Qur’anic text in comparison with their target equivalents.

4.1. Amphibological syntax

Syntactic ambiguity in this study is limited to a:ya:t at-ta^ca:nuq (‘inter-embracing verses’) which have not been duly investigated from a syntactic perspective in a contrastive translational context. That the double inter-embracing pause signs (^{∴ ∴}) are superscripted only in the Qur’anic Arabic verses necessitates quoting them first in transliterations to be followed by ST glossings and by their TT translations. The verses in focus are analysed and discussed in their numerical order in the ST and accordingly in the TT.

4.1.1. PP attachment

Example (1)

Verse: dha:lik al-kita:b la: rayb fi:h huda: li-l-muttaqi:n.

Gloss: That the-Book no doubt in-it guidance to-the-pious.

Source. (ENA, September 5, 2022)⁴

In this verse and the subsequent ones, the superscripted pause signs (^{∴ ∴}) cause the ST verses to be interpreted in two ways and thus manifest amphibological ambiguity. In the verse above, syntactic ambiguity develops from an NP attachment or a PP attachment based on which pause sign the reader chooses to stop at. Pausing at the first sign (^∴) generates the interpretation “That Book no doubt [in it a guidance for the pious]” which results from a PP attachment. Pausing at the second sign (^∴) generates a different interpretation “That Book [no doubt in it] a guidance for the pious” which develops from an NP attachment. Figure 2 disambiguates this verse with tree-diagrams.

Figure 2. Disambiguation of verse (2:2) with tree-diagrams

Two different pauses lead to two different readings, which give rise to two different interpretations and accordingly two different translations based on whether the PP fi:h (‘in it’) is post-positionally attached to the preceding NP or pre-positionally attached to the following NP (Az-Zamakhshari: 1998 [P1]: 145, Ibn ^CA:shu:r 1984 [P1]: 222–223). Figure 3 provides a syntactic treebank of the verse.

Figure 3. Syntactic Treebanking of Verse (2:2)
Source. (ENA, September 5, 2022)⁵

Figure 3 depicts the post-positioning of the PP fi:h as the predicate of the subject of the negative particle la: (‘no’) whereas its pre-positioning is not depicted. Although the semantic duplicity of this amphibolous verse is potentiated by Qur’an exegetes, each translator opts for one meaning, excludes the other, and thus detracts from the information structure of the ST. SI, MP, HS, MS, MK, and JA choose to pause at the PP fi:h (‘in it’), relaying a holistic meaning that predicatively or appositively describes that Book, the Qur’an, as being a guidance for the pious. YA chooses to pause at the NP la: rayb (‘no doubt’), transferring a partitive meaning which describes the Qur’an as including in a number of its verses guidance for the pious. One informational chunk is intratextualised, but the other is not intratextualised or paratextualised although it is of equal note in Qur’anic exegeses (see Al-Ali & Al-Zoubi 2009: 231).

Example (2)

Verse: min al-ladhi:n qa:lu: a:manna: bi-afwa:hihim wa-lam tuʔmin qulu:buhum wa-min al-ladhi:n ha:du samma:^cu:n li-l-kadhib.

Gloss: from who said believed-we with-mouths-their and-not believed hearts-their and-from who Judaised listeners to-the-falsehood.

Source. (ENA, September 5, 2022)⁶

In a similar vein, verse (5:41) is syntactically ambiguous owing to the PP min al-ladhi:n ha:du: (‘of those who Judaised’), which is attachable either to the preceding PP min al-ladhi:n qa:lu: a:manna: bi-afwa:hihim wa-lam tuʔmin qulu:buhum (‘of those who say “we believe” with their mouths but their hearts do not believe’) or to the following NP samma:^cu:n li-l-kadhib (‘listeners to falsehood’). Figure 4 resolves this amphiboly with tree-diagrams.

The potentiality of both meanings hinges upon where recitation stops, at the first pause sign or at the second. Both meanings are communicated by Az-Zamakhshari: (1998 [P2]: 235) whereas only the latter meaning features in Ibn ^CA:shu:r (1984 [P6]: 198) who prefers to pause at the PP min al-ladhi:n ha:du: (‘of those who Judaised’). Figure 5 sketches this latter meaning.

Figure 4. Disambiguation of verse (5:41) with tree-diagrams

Figure 5. Syntactic Treebanking of Verse (5:41)
Source. (ENA, September 5, 2022)⁷

As Figure 5 shows, the language group of the University of Leeds diagrams only one meaning which is favoured only by Ibn ^CA:shu:r (1984). The former meaning transpires in SI, MP, MS and MK while the latter meaning occurs in YA, HS and JA. In each translation, as shown, one meaning is unjustifiably chosen over the other–a practice subtracting from the semantic duplicity and propositional content of the Qur’anic verse.

4.1.2. VP attachment

Example (3)

Verse: wa-la: tulqu: bi-aydi:kum ila: at-tahluka wa-ahsinu: inna alla:h yuhibb al-muhsini:n.

Gloss: and-not throw by-hands-your to danger and-do well-you indeed God loves the-right doers.

Source. (ENA, September 5, 2022)⁸

In this verse, syntactic ambiguity results from a VP attachment, whereby the imperative VP wa-ahsinu: (‘and do good’) is attributable either to the preceding negative imperative VP wa-la: tulqu: bi-aydi:kum ila: at-tahluka (‘and do not put your head into the lion’s mouth’) or to the following causal inna-sentence inna alla:h yuhibb al-muhsini:n (‘God does love the doers of good’). Both interpretations depend upon which pause sign the reciter halts at. Az-Zamakhshari: (1998 [P1]: 397) decides to pause at the third VP and therefore post-positionally attaches it to the preceding VP whereas Ibn ^CA:shu:r 1984 [P1]: 222–223) takes a neutral stance from these two mutually exclusive pause signs. Figure 6 resolves the amphibology of the verse and depicts its two readings in tree-diagrams.

Figure 6. Disambiguation of verse (2:195) with tree-diagrams

The first tree-diagram syndetically conjoins the VPs wa-anfiqu: (‘and spend’), wa-la: tulqu: bi-aydi:kum ila: at-tahluka (‘and do not put your head into the lion’s mouth’) and wa-ahsinu: (‘and do good’). The second tree-diagram paratactically appends the third VP to the following causative inna-sentence. Figure 7 gives a syntactic treebank of the former meaning.

Figure 7. Syntactic Treebanking of Verse (2:195)
Source. (ENA, September 5, 2022)⁹

The treebank in Figure 7 coincides with the former meaning depicted by the first tree-diagram in Figure 6. Either meaning, but not both, figures in the given translations; one at the expense of the other. The former meaning occurs in HS, MK and JA; the latter in SI, MP, YA and MS. The exegetes do more justice and are more faithful to the Qur’an than the translators because the former always introduce a comprehensive pool of interpretations whereas the latter choose one interpretation from this pool over the others, often without paratexts. The causality of inna-sentence, which figures in the parsing of the Qur’an (see Ad-Darwi:sh 1980 [V1]: 285, Daas 2004: 24), is only attended by YA in his use of the coordinating conjunction ‘for’, which establishes a causal relation between the third VP wa-ahsinu: (‘and do good’) and the inna-sentence inna alla:h yuhibb al-muhsini:n (‘for Allah loveth those who do good’).

Example (4)

Verse: wa-ashhadahum ^cala: anfusihim a-last bi-rabbikum qa:lu: bala: shahidna: an taqu:lu: yawm al-qiya:ma inna: kunna: ^can ha:dha: gha:fili:n.

Gloss: and-testified-them upon selves-them am-not-I by-Lord-your said-they yes testified-we that say-you day-resurrection indeed were-we of this unaware.

Source. (ENA, September 5, 2022)¹⁰

Likewise, the verse above logs the same aspect of structural ambiguity as a result of VP attachment, whereby a recitational pause before or after the VP shahidna: (‘we testify’) creates two readings. One reading is wa-ashhadahum ^cala: anfusihim a-last bi-rabbikum qa:lu: bala: shahidna: (‘He made them testify against themselves: am I not your Lord? They said: Yes (you are), we testify’)–a reading conditioned by the pause at the second sign, i.e., right after the VP. The other reading is shahidna: an taqu:lu: yawm al-qiya:ma inna: kunna: ^can ha:dha: gha:fili:n (‘We testify that you might say on the Day of Resurrection that we were unaware of this’)–a reading provided by the pause at the first sign, i.e., right before the VP. Az-Zamakhshari: (1998 [P2]: 530) favours the former reading while Ibn ^CA:shu:r 1984 [P9]: 168–169) considers both readings possible, as the tree-diagrams in Figure 8 shows.

Figure 8. Disambiguation of verse (7:172) through tree-diagrams

Ad-Darwi:sh (1980 [V3]: 492) supports the second reading in parsing the CP of the third VP as maf^cu:l li-ajlih (‘object of cause or reason’), and thus agrees with the treebanked reading that describes the CP as an SC by form and accusative of purpose by function. Daas (2004: 207) considers this CP as a resumptive statement uttered either by Adam’s posterity or by the angels. Figure 9 shows the syntactic treebanks drawn by the Language Research Group.

Figure 9. Syntactic Treebanking of Verse (7:172)
Source. (ENA, September 5, 2022)¹¹

However, the Language Research Group members choose to stop at the second pause sign right after the VP, which they regard as a verbal sentence (VS) by form and direct object (DO) of the speech verb qa:lu: (‘said’) by function. This reading appears in SI, MP, YA, HS, MK and JA; the other reading emerges only in MS.

4.1.3. AdvP attachment

Example (5)

Verse: fa-innaha: muharrama ^calayhim arba^ci:n sana yati:hu:n fi: al-ard.

Gloss: then-indeed-it forbidden on-them forty years wander-they in-the-land.

Source. (ENA, September 5, 2022)¹²

In the verse above, the grammatical ambiguity comes from an AdvP attachment, an NP functioning as an adverb(ial) or adjunct of time (see Simpson 2004: 10). AdvP attachment logs a case of amphibology because different interpretations are equally possible based upon where the pause is made (see Al-Ali & Al-Zoubi 2009: 235). The optionality and mobility of adjuncts or adverbials bear a few nuances of meaning therewith. According to Qur’anic parsing (see Ad-Darwi:sh 1980 [V2]: 449, Daas 2004: 140) or exegetes (see Az-Zamakhshari: (1998 [P1]: 223), the AdvP arba^ci:n sana (‘forty years’) is ascribable either to muharrama (‘forbidden’), the predicate of inna-sentence, thus meaning “the land shall be forbidden for them for forty years”, or to yati:hu:n fi:-l-ard (‘they shall be wandering through the land’), the circumstantial clause, thus meaning “straying thereabouts for forty years”. Figure 10 resolves this amphibology and illustrates both readings through tree-diagrams.

Figure 10. Disambiguation of verse (5:26) with tree-diagrams

As shown in Figure 10, there are two possible readings of the verse above, which answer the question ‘How long shall they (i.e., the Israelites) be forbidden from entering the Terra Sancta (‘the Sacred Land’) or how long shall they be wandering in the wilderness?’ The answer is ‘forty years’. Figure 11 illustrates the treebank of this verse.

Figure 11. Syntactic Treebanking of Verse (5:26)
Source. (ENA, September 5, 2022)¹³

Figure 11 places the syntactic pause right after the AdvP and thus features just the former meaning depicted in the first tree-diagram in Figure 10 and explicitly rendered by YA, HS and MK. The latter meaning is explicit in SI and MS but implicit in MP and JA. Both target meanings are totally dependent upon whether the recitational pause takes place before the time adjunct or after it. Whether forbidden or bewildered, the Israelites received this punishment down from God in reply to Moses’s supplication for help against their obstinacy.

4.1.4. RC attachment

Example (6)

Verse: qawm nu:h wa-^ca:d wa-thamu:d wa-l-ladhi:n min ba^cdihim la: ya^clamuhum illa: alla:h.

Gloss: community Noah and-Ad and-Thamud and-who from after-them not know-them except God.

Source. (ENA, September 5, 2022)¹⁴

Verse (14:9) carries two possible interpretations which differ according to which pause sign the reader opts to stop at. The RC wa-l-ladhi:n min ba^cdihim (‘those who came after them’) can be attached backward to the NP wa-thamu:d (‘and Thamudites’), (i.e., ‘those who came after Noahites, Adites, and Thamudites.’), or forward to the VS la: ya^clamuhum illa: alla:h (‘are known only to God’). Figure 12 resolves this amphiboly with tree-diagrams.

Figure 12. Disambiguation of verse (14:9) through tree-diagrams

The verse is diagrammatically interpreted in two different ways based on where the RC is attached: regressively or progressively. One possible interpretation is ‘Have you not received the news of those before you: Noahites, Adites, Thamudites, and those after them?’ The other interpretation is ‘And those after them are known only to God.’ Both interpretations feature in Az-Zamakhshari: (1998 [P3]: 365) while only the second reading occurs in Ibn ^CA:shu:r (1984 [P13]: 196) and Ad-Darwi:sh (1980 [V5]: 162–163). Surprisingly, the first reading transpires in all the seven translations to the exclusion of the second which is exegetically expounded.

4.2. Anacoluthic syntax

Anacoluthon is said to be a rhetorical stylistic feature of the Qur’anic discourse (Justice 1987, Darir 2012) as it is of the Biblical discourse (Schipper 2012), and the literary discourse (Tüfekçican 2017, Rangarajan 2017, among some others). Due to space and word restrictions, a few representative cases of anacoluthon are discussed in the following lines, as the purpose is exemplificative rather than exhaustive. However plethoric, anacoluthic examples are typified and presented below in order of importance and preponderance.

4.2.1. Anapodoton

The most common type of anacoluthic syntax in the Qur’an is anapodoton, which is shaped by the lack of an apodosis (a main clause) in a protatic (conditional) sentence.

Example (7)

Verse: wa-law ann qur?a:n suyyirat bih al-jiba:l aw qutti^cat bih al-ard aw kullim bih al-mawta: bal li-l-a:h al-amr jami:^ca:.

Gloss: and-if indeed a Qur’an be-moved by-it the-mountains or be-cracked
by-it the-earth or be-addressed by-it the-dead rather to-God the-matter all.

Source. (ENA, September 5, 2022)¹⁵

The verse above is a typical example of anapodotonic anacoluthon that signals the absence of jawa:b ash-shart (‘the apodosis’) as a consequence of fi^cl ash-shart (‘the protasis’). This hiatus is easy to fill from the context (of situation) with some measure of subjectivity, however. This is exactly what has been done in brackets or parentheses in each translation, in which the apodosis reads as “it would be this Qur’an” (SI), “this Qur’an would have done so” (MP), “this would be the one!” (YA), and “it would not have been other than this Quran” (MK). HS has kept the main clause implicit and rendered the anacoluthon into his translation. MS has recontextualised the apodosis differently as “the unbelievers still would not believe.” JA has transferred and neutralised the protasis as it is without explicitation. The explicitation of the apodosis as “it would be the Qur’an” is agreed on by a majority opinion (see Az-Zamakhshari: (1998 [P3]: 352, Ibn ^CA:shu:r (1984 [P13]: 143) and by parsing consensus (see Ad-Darwi:sh (1980 [V5]: 123, Daas 2004: 301).

4.2.2. Anantapodoton

This is the second form of anacoluthon, a subcategory of anapodoton, in which the sentence trails off meaningfully and leaves the subordinate clause uncompleted, without a main or superordinate clause to complete its meaning.

Example (8)

Verse: wa-idha: qi:l la-hum ittaqu: ma: bayn aydi:kum wa-ma: khalfakum la-^callakum turhamu:n.

Gloss: and-when be-said to-them fear what between hands-your and-what behind-you perhaps-you be-pitied.

Source. (ENA, September 5, 2022)¹⁶

This verse showcases anantapodotonic anacoluthon–a case of hypotaxis in which the syntactic subordination is left unfinished and the subordinate clause is left uncompleted by a superordinate clause. On subordination, Simpson (2004: 61) stipulates that a subordinating conjunction such as ‘when’ is used to append a subordinate clause to a main clause in order to communicate a complete meaning, as diagrammatically conceptualised in Figure 13.

Figure 13. A subordinate clause appended and not appended to its main clause

The conceptual nexus of this subordinate relationship conjures up an image of a box (a subordinate clause) leaning on another supporting box (a main clause) which, if pulled away, causes the leaning one to topple (see Simpson 2004). Figure 13 illustrates a sound hypotaxis in the first image and an anacoluthic hypotaxis in the second. Translators of the verse under scrutiny are assigned an ineludible responsibility to recover from the context (of situation) a main clause and append it to the subordinate clause. At this juncture, explicitation is an ineluctable strategy. Exegetically, Az-Zamakhshari: (1998 [P5]: 181) and Ibn ^CA:shu:r (1984 [P23]: 31) make an inference that the main clause a^cradu: (‘they ignore’) is inferable from the following co-textual verse (36:46). Translationally, this anantapodotonic anacoluthon is maintained in all the given translations except in those by MP and YA who smartly tended to this anacoluthic syntax and parenthesised a meaningful complement.

4.2.3. Particula pendens

Another guise of anacoluthon is particula pendens which relates to correlative conjunctions (e.g. imma:…aw ‘either…or’)–a common case in which only the first particle of a correlative expression representative of a binary choice or opposition is mentioned (see Greene et al. 2012: 46, Greene & Cushman 2016: 11).

Example (9)

Verse: fa-imma: nadhhabann bik fa-inna: min-hum muntaqimu:n.

Gloss: so-either we-wend with-you then-we from-them revenging.

Source. (ENA, September 5, 2022)¹⁷

In this anacoluthic verse, the second particle of a correlative conjunction, which is frequently used in the Qur’an to denote a binarised option, i.e., a choice from two alternatives, imma:…wa-imma: or imma:…aww (‘either…or’), is apparently lacking and therefore throws down a massive challenge to Qur’an translators. The first correlative member imma: (‘either’) occurs in the SL text while its complementary fellow, aww (‘or’), is intraversially absent but interversially present (43:41 & 43:42). Az-Zamakhshari: (1998 [P5]: 445) interprets imma: as a conditional ‘if’ and considers it a protatic rather than coordinate structure the meaning of which is ‘If we decree your death, we are going to wreak vengeance on them’ and ‘If we want to show you the torment we promised them, we have the full power over them to do so.’ This reading is exegetically mirrored by Ibn ^CA:shu:r (1984 [P25]: 217–218) and is syntactically parsed by Ad-Darwi:sh (1980 [V9]: 89–90) and Daas (2004: 529). Despite the consensus on the protatic-apodotic structure of this verse, the Qur’an translators interpreted and rendered it differently: explicitly correlatively as ‘whether…or’ (SI and JA) and ‘either…or’ (MS), concessively as ‘even if’ (YA and MK), and compliantly conditionally as ‘if’ (MP and HS).

4.2.4. Aposiopesis

Often associated with anacoluthon is aposiopesis occurring when a sentence breaks off not to continue (see Mieszkowski 2009: 648) and defined as a pause that sometimes speaks eloquence (see Langley 1835: 57). Bussmann (1996: 74) regards it as a rhetorical trope which shortens an expression with a breakoff to express an alarm or concern and the unexpressed thought of which is easily perceivable. Related to aposiopesis is nominativus pendens in which a sentence is begun with an apparently predicateless subject (see Dupriez 1991: 35).

Example (10)

Verse: inn al-ladhi:n kafaru: bi-dh-dhikr lamma: ja:ʔahum wa-innahu la-kita:b ^cazi:z.

Gloss: indeed who disbelieved in-the-Qur’an when it reached-them and-indeed-it certainly-scripture unassailable.

Source. (ENA, September 5, 2022)¹⁸

In the verse above, anacoluthon figures aposiopesically in an inna-sentence in which the emphatic or assertive particle inna (‘indeed, certainly’) heads a nominal sentence consisting of a subject in the accusative case and a predicate in the nominative (Abu-Chacra 2007: 193). In this verse, the accusative subject is present whereas its nominative predicate is absent and is left for the readers and translators to figure it out. Exegetically, Ibn ^CA:shu:r (1984 [P25]: 307) asserts the ellipsis of inna-predicate and its recoverability from the context (of situation), e.g., as “they have lost this life and the afterlife.” The exegetical interpretation of the verse is further supported by the syntactic parsing undertaken by Ad-Darwi:sh (1980 [V8]: 569) who infers the ellipted predicate from the preceding co-text (41:40) and interprets it as “are not hidden from us.” Translationally, some translators suggested few empty-slot fillings, as in “are guilty” (MP), “are not hidden from Us” (YA), “do not know” (MS), and “shall receive the punishment” (MK), whereas the remaining translators (SI, HS and JA) maintained the aposiopesic anacoluthon in their translations.

5. Discussion

The Qur’an abounds in multileveled ambiguous and multifaceted elliptical structures attesting its idiosyncratic rhetorical style and sometimes resisting formal correspondence and dynamic/functional equivalence between Arabic and English. Although previous translation studies on Qur’anic ambiguity and ellipsis are manifold, there is a manifest paucity of past literature on amphibol(og)y and an extreme dearth of research into anacoluthon in Arabic in general and Qur’anic Arabic in particular. Therefore, the need for the present study has arisen to conduct a seminal investigation of the problematics of rendering amphibolous and anacoluthic syntactic structures from Qur’anic Arabic into English to examine and assess the translational quality of transposing meaning duplicity and interpretive multiplicity (multi-interpretability) from the SLT to the TLT.

The twofold purpose of the study has been to explore how Qur’an translators resolved the amphibologies arising from the pause signs (∴ ∴) and how they sequentialised the anacolutha (…) arising from ellipted catenae (ellipses easy to retrieve and conceive from a pre-text, a co-text, an intra-text, an epi-text or a con-text). Regarding amphibology resolution, the seven Qur’an translators in comparative and contrastive focus differ vastly on which pause sign to stop at. They rendered one single meaning and excluded another which is equally potentiated by Qur’an exegetes. Their inclusion of a single reading and exclusion of the other detracted from the propositional content of the deep structure of the amphibolous or amphibological SL verses. The surface meaning of the syntactically ambiguous verses borne with duplicities of meaning has been transposed with a miscellany of categorical shifts (Catford 1965 qtd. in Najjar et al. 2019), but at the expense of deep meaning. Implicitation, “where a given target text is less explicit (more implicit) than the corresponding source text” (Becher 2011: 19), is used by the Qur’an translators in focus for rendering amphibolous Qur’anic syntax. The corresponding target translation (text) is less explicit than the source text as the translators are unable to imitate such an inimitable Qur’an-specific phenomenon. Paratextualization is a translation procedure highly (re)commended by translation theorists for resolving ambiguous structures (e.g., Newmark 1982, 1988, Munday 2008, among others).     

Concerning anacoluthon sequentialisation, Qur’an translators have introduced hetero-subjective stances and inadequate interpretations. In all anacoluthic verses, ellipses, apodotic or hypotactic, have been transposed into the TLT with varying elliptical-gap fillings. It is quite easy to point out the interpretive variations among the translators in parenthesising the elided materials. Parenthesisation is an intratextual translational technique employed by some of the translators in comparison to reproduce a sound reader-friendly grammatical structure in TLT. Explicitation, “[a] stylistic translation technique which consists of making explicit in the target language what remains implicit in the source language because it is apparent from either the context or the situation.” (Vinay & Darblenet 1995: 342, qtd. in Becher 2011: 17), is used as a procedure for restructuring anacoluthic verses into English. The corresponding TT is rendered more explicit than the ST (see Becher 2011: 19). Other translators took neutral stances, with neutralisation as intermediary between explicitation and implicitation, and transferred unsound syntactic structures in English.

Both ST (i.e., the Qur’an)-oriented translators and TT (i.e., the translation)-oriented translators seem to have been motivated by a skopos, a purpose, to achieve for reasons of translation quality assessment parameters from either side, such as acceptability (subscription to TL norms), adequacy (subscription to SL norms) and accuracy (subscription to ST content). Such parameters of (Qur’an) translation quality assessment (see, Toury 1995) are instigated, I tend to claim, by antinomies of fidelity and readability, faithfulness and reader-friendliness–two dilemmatic issues still difficult to resolve or reconcile. Adequacy and accuracy of both amphibology resolution (disambiguation) and anacoluthon sequentialisation necessitate the consideration of what Salama (2021) recommends as further extension of Genette’s (1997) concepts of ‘paratexts’ and ‘epitexts’, i.e., paratextual materials/references appended or not appended to the translated text, respectively. 

6. Conclusion

It is exegetically and translationally assumed that waqf al-mu^ca:naqa (‘inter-embracing pause’) in the Qur’an is part of the problematics of Qur’anic interpretation and translation (see Al-Ali & Al-Zoubi 2009, ^cImra:n 2018), and so is fuqda:n at-tata:bu^c (‘anacoluthon’). This study has taken the initiative to test this assumption in a contrastive translational context and prove these two syntactically rhetorical phenomena to be challenging to the Qur’an translators. As regards the former, the reciters consciously apply prosodic disambiguation to the syntactic ambiguity of the verses while Qur’an translators find themselves unable to do so and are left with the option of choosing one interpretation over the other. Variational recitations beget variational interpretations and accordingly translations–a conclusion in line with Al-Ali & Al-Zoubi (2009: 235) who contend that “different attachments lead to different interpretations.” As concerns the latter, some of the translators are ST-oriented for reasons of faithfulness while the others are
TT-oriented for reader-friendliness purposes. Each translatorial orientation has brought with it a caveat: (a) blind faithfulness to the SLT has engendered sentential fragments and ungrammatical clauses into the TLT when it relates to anacolutha, and (b) excessive reader-friendliness has resulted in hermeneutical gaps and epistemological voids, which might have been filled by compensatory paratextual strategies, as in footnotes or endnotes (see Newmark 1988: 220, Munday 2016: 129, Abdelaal 2019: 3).

This study argues that anacoluthon and amphiboly are subsumable under translating the implicit (Darir 2012) and the invisible (Al-Kharabsheh & Al-Azzam 2008), respectively. Qur’an translators vary in their unidirectional translation practice and add to Klaudy’s (2005) explicitation-implicitation dyad a demarcation zone I am prone to call ‘neutralisation’. There are clear cases in which translators act as explicitators who apply adjustment-by-addition to the source text. Anacoluthon is found by Darir (2012) to be a guise of explicitation according to which the TT is more explicit than the ST, a finding in line with Bisiada (2016) who proves that sentence splitting is an explicitating strategy rather than a process triggered by the target language. This finding is strongly supported by Raoufkazemi et al. (2020) who revealed that the texts rendered by experienced translators are more explicit than the original.

There are cases in which translators serve as implicitators who apply adjustment-by-subtraction (and detraction) in the source text. There are also cases in which translators take a neutral stance from the source text, neither adding to nor subtracting from it. As for the thesis at hand, some translators explicitated anacoluthic structures while others implicitated and neutralised the ambiguous structures. This act gives rise to a triadic frame categorising Qur’an translators as cross-cultural explicitators, implicitators, and neutralisers.  

No sooner does it seem that the voice of the Qur’an goes or trails off on tangents than it gets clear that no digressional remarks have been made to hark and circle back to the topics covered. There are notable individual differences in the resolution of syntactic ambiguity and sequentialisation of anacolutha in Qur’anic translations because of the plausible variations in translatorial inferences of the rhetorically amphibological and anacoluthic Qur’anic syntax. Amphibologies are contextually resolvable and anacolutha are cotextually sequentialisable. The context (of situation) plays a pivotal role in resolving amphiboly and completing anacoluthon (see MacDonald, Pearlmutter & Seidenberg 1994, Bousquet, Swaab & Long 2019). Although Qur’an translators have latitude in resolving amphibology and sequentialising anacoluthon, they epitomise heterosubjectivity in their interpretive choices and translational preferences often independent of exegeses–asymmetricity at large. Such asymmetricity in religious translation (‘alterity’ elsewhere, see Makutoane, Miller-Naudé & Naudé 2015) is not specific to the Qur’an but reverberates to the Bible, as well. The fact that “translation is a prime player in intercultural communication” (House 2019: 3), and equally in interreligious dialogue, necessitates striving for inter-objectivity and inter-symmetricity in religious translation in order not to mistranslate and misrepresent the divine message. “Translation-cum-faith must always examine the context of its production and be attuned to the context of its reception, willing to make adjustments in order to best communicate its message” (Blumczynski 2017: 89).

Appendix 1. Abbreviations and typographical conventions

Appendix 2. Transliteration symbols for Arabic vowels and consonants

Source. Retrieved and adapted from http://www.ijaes.net/Author/Help and accessed on 07/03/2020.

^{1 See Appendix 1 for full forms of the abbreviations and typographical conventions and Appendix 2 for transliteration symbols for Arabic vowels and consonants.}

^{2 For an informative review of the untranslatability of Qur’anic discourse due to its unique and sophisticated character, see Abdul-Raof (1999) and Hassanein (2017).}

^{3 This corpus has been designed by the Language Research Group at the University of Leeds. For a counter-reference, see http://corpus.quran.com/translation.jsp.}

^{4 http://corpus.quran.com/translation.jsp?chapter=2&verse=2}

^{5 http://corpus.quran.com/treebank.jsp?chapter=2&verse=2&token=0}

^{6 http://corpus.quran.com/translation.jsp?chapter=5&verse=41}

^{7 http://corpus.quran.com/treebank.jsp?chapter=5&verse=41&token=17}

^{8 http://corpus.quran.com/translation.jsp?chapter=2&verse=195}

^{9 http://corpus.quran.com/treebank.jsp?chapter=2&verse=195&token=10}

^{10 http://corpus.quran.com/translation.jsp?chapter=7&verse=172}

^{11 http://corpus.quran.com/treebank.jsp?chapter=7&verse=172&token=15}

^{12 http://corpus.quran.com/translation.jsp?chapter=5&verse=26}

^{13 http://corpus.quran.com/treebank.jsp?chapter=5&verse=26&token=0}

^{14 http://corpus.quran.com/translation.jsp?chapter=14&verse=9}

^{15 http://corpus.quran.com/translation.jsp?chapter=13&verse=31}

^{16 http://corpus.quran.com/translation.jsp?chapter=36&verse=45}

^{17 http://corpus.quran.com/translation.jsp?chapter=43&verse=41}

^{18 http://corpus.quran.com/translation.jsp?chapter=41&verse=41}

About the authors

Hamada Hassanein

Author for correspondence.
Email: h.hassanein@psau.edu.sa
ORCID iD: 0000-0003-3770-6405

Associate Professor

References