Disruptions in Translational Intertextuality across Multilingual Translations of a Historical Inscription

This is the first research study to investigate translational intertextuality across multilingual translations of a historical inscription through retranslation and relay translation. It aims to offer a translational chain by linking the studied translations that involve different translational strategies and reveals intra-and multi-lingual alterations within the collection of the analyzed works. The corpus of the study consists of six academic works – on the interlingual translations of a historical inscription in Alanya, Türkiye – in five different languages, including the source language. To analyze the corpus, qualitative content analysis and purposeful sampling are applied. It is revealed that the translations tend to change their textuality through semantic alterations and application of translational phenomena, and thanks to the alterations, it is seen that the source text evolved through translations. It is suggested that translational intertextuality has a crucial role in the creation of new texts in a target language based on a translation, and the studied corpus illustrated that any disruptions in translational intertextuality affect the other rings of the translation chain, directly or indirectly helping to disseminate false knowledge or/and foreignizing the resultant text for the target audience. Moreover, it is highlighted that an ulterior translation can be affected by these disruptions; therefore, the disrupted prospective translational intertextuality could influence the intertextual ties across disciplines established by translations.


Introduction
This is the first paper handling the impacts of relay translation and retranslation upon translational intertextuality in historical texts by linking the renditions through comparative analysis to create a translational chain that involves different translational strategies and indicates the intra-and multilingual alterations.Translational intertextuality highlights coexistent renditions' ability to affect a current and ulterior translation through incorporating connections between anterior and posterior renditions.The present paper aims to observe semantic alterations and translational blurs in a translation chain -formed through relay translation and retranslation, -by employing qualitative content analysis and purposeful sampling in a corpus of six different academic works.
The paper firstly elaborates on the interrelation between relay translation, retranslation, domestication and foreignization, and translational intertextuality.Secondly, it justifies the authors' selection of qualitative content analysis and purposeful sampling.Thirdly, a corpus of six different academic works is presented, their comparative analysis is conducted, and the highlighted translational differences are discussed.Lastly, the study's significance is highlighted, and then the paper is concluded with final remarks.

Relay Translation
Relay translation is a translational process whereby a new translation is rendered based on a mediatory text.In this case, the mediatory rendition is called a relay, and the output, the subsequent translation, is referred to as a relayed translation.For the ultimate text to be called a relayed translation, the mediating text must be produced from an initial source text to be received by an authentic audience in the target culture.It also applies to a translation that will serve as a relay for prospective (relayed) translations.Therefore, a temler ve diliçi ve dillerarasındaki değişimler öne çıkarılmıştır.Çalışmanın bütüncesi altı akademik çalışma ve kaynak dil dahil olmak üzere beş dil içermektedir.Bahsedilen altı çalışma Türkiye'nin Alanya ilçesinde bulunan bir tarihi kitabenin dillerarası çevirilerinden oluşmaktadır.Bütüncenin çözümlemesinde nitel içerik çözümlemesi ve nitel amaçlı örneklem yöntemleri kullanılmıştır.Çözümleme sonucunda, anlambilimsel değişikliklere ve çeviri görüngülerimin uygulanmasına maruz kalan çevirilerin metinliklerini değiştirmeye yatkın olduğu görülmüştür.Ayrıca, bu yapılan değişiklikler sonucunda kaynak metne yeni anlamlar kazandırıldığı gözlemlenmiştir.Bu bağlamda, çevirisel metinlerarasılık kavramının hedef dilde oluşturulacak çeviriye dayalı çeviri metinlerde önem arz ettiği vurgulanmıştır.Buna ek olarak, çevirisel metinlerarasılıkta muhtemel bozulmaların çeviri zincirinin diğer halkalarını da etkilediği izlenmiştir.Bu bozulmalar doğrudan veya dolaylı olarak yanlış bilginin yayılmasına ve/ya da hedef dildeki okuyucuya yabancılaştırma hissi vereceği gözlemlenmiştir.Son olarak, gelecekte yapılacak çevirilerin bu bozulmalardan etkileneceği vurgulanmış ve böylelikle bozulmuş ileri dönük çevirisel metinlerarasılık bağlarından çeviriler aracılığıyla disiplinler arası oluşturulmuş metinlerarası bağlarının da etkilenebileceği görülmüştür.. Anahtar Kelimeler: Çevirisel metinlerarasılık, aktarım çevirisi, yeniden çeviri, tarihi kitabe, bozulmuş ileri dönük metinlerarasılık relayed translation can include a plethora of translational strategies thanks to the source and mediatory translation's aim to reach a genuine audience.The systematization of a relay and relayed translation are discussed in this part.Dollerup (2014: 3) describes that a relayed translation relies on "a translation that has a genuine audience in the first target language" and thus creates a translational network of at least three languages as follows: "source text (L1), the first translation [the relay] (L2), and the relayed translation (L3)" (Dollerup 2014: 3).Moreover, it can be added that there are theoretically an infinite number of interlingual renditions from a relay (Dollerup 2014: 3).Pym (2011: 89) features that this translation technique is essentially applied when the source text is from a "lesser-known language".To exemplify, André (2003: 62) states that not having enough skilled translators to conduct translations from 'periphery' languages to canonized languages was common in the 18 th and 19 th centuries.Gambier (2003: 62) goes back further and mentions the macaronic translations of the Toledo School during the 12 th and 13 th centuries from Arabic and Hebrew.Currently, English as the lingua franca assumes the mediating role with its more than 1.5 billion people 1 speakers.In the case of translations from lingua franca, multiple translations can co-occur due to the language's ubiquity.
Relay translation should not be confused with indirect translation, in which the initial translation is not intended for an authentic audience but just serves as a mediatory copula between the first and third or later translations (Dollerup, 2014: 3).Indirect translation incorporates a mediatory translation and therefore at least three languages are at play in the rendition of the ultimate translation.Yet unlike in relayed translation, the mediating translation is not created for a genuine reader; it is produced only for the sake of the ultimate translation into a third language.
To sum up, a relayed translation is assumed to be read by a genuine audience in the target language and culture.However, indirect translation does not aim for a real audience, and it serves as an interlingua between two different languages.Because the intermediatory translated texts in the present corpus are intended for real readers, relay(ed) translation was operationalized to refer to these texts.

Retranslation
Translation is presumably a product of intertextual and interlinguistic transfer procedure whereby two or more linguistic systems are mediated to allow the readership to be informed about each other.It can be understood that this intertextual and interlingual, sometimes intralingual, practice is language-and culture-bound.Yet, thanks to this boundness, a source text may need to be translated again in view of source texts' change 1 https://www.weforum.org/agenda/2019/11/countries-that-speak-english-as-a-second-language/(English).
in time thanks to a number of interventions, "political reasons" or languages' unstable, everchanging nature (Koskinen and Paloposki 2010: 294).This translational phenomenon is called retranslation.
Retranslation, as Washbourne (2013: 608) calls "second-hand translations, second-generation translations", can be defined as "the new partial or complete translation of a translated text into the same language" 2 (Gambier 1994: 413).Koskinen and Paloposki (2010: 294) classify retranslation as "second or later translation of a single source text into the same target language".Furthermore, Şahin et al. (2019: 166) broaden retranslation's concept by highlighting that "every translation after the first translation of the source text, regardless of the language it is translated into, is a retranslation".Şahin et al. (2019: 166) rationalize their argument by underlining that a first rendition may be embodied in the target language as a source text since it is a clone.Therefore, it can be observed that retranslation refers to not only the second or later translations of a source text into the same language but also every rendition of a single source text into various languages.
To the authors' best knowledge, the central moral thesis of retranslation can be defined with labile conditions of language, changes in "social context and the evolution of translation norms" (Tahir-Gürçağlar 2020: 485) affecting some source texts, "language change", and economic interests (Dollerup 2014: 2).From another perspective, the rationale behind the fact that the first translations may not be a proper translational rendering is that the initial translation features a dilemma, i.e., whether to be accepted within the target culture or not (Desmidt 2009: 671).As noted above, the first translations tend to use the domestication technique to abolish cultural boundaries; however, in time, the target culture "allows for and demands new translations" (Desmidt 2009: 671) -retranslations relatively less domesticated and more foreignized -in the case of acceptance of the first translation.Moreover, Lowe (2014: 415) mentions the extant translations' possible "lacking literary quality" and a text's or author's "voice" can be represented thanks to a new rendition.Therefore, as inferred from the above discussion, the definition of retranslation can be broadened to incorporate a translational shift from being target-orientedness in the event of the initial translation's acceptance by the target culture that may also request a new translation relying on foreignization.

Domestication and Foreignization
"Domestication" and "foreignization" are two translational strategies offered by Venuti (2008).These concepts illustrate different approaches so as to establish a link(s) between two distinct languages and cultures in a translational process.Nevertheless, Venuti (2008: 19) underlines that these strategies should not be regarded as "binary oppositions" yet "ethical attitudes".The former, as Venuti (2008: 19) also calls "fluency", aims for an "ethnocentric reduction of the foreign text to dominant cultural values in English" (Venuti 2008: 68).The latter, which is referred to as "resistancy" by Venuti (2008: 19), can be considered an upheaval against ethnocentrism (Venuti 2008: 68) and it
can be stated that it attempts to preserve a source text's essence and authenticity in a target language and culture through borrowing and calque.
In the corpus of this research paper, it is noticed that both strategies are used in the inscription's interlingual translations.Moreover, several translators prefer to use the strategies at the lexicological and phraseological level paving the way for establishing macaronic texts for a target reader, in which several parts of the inscription are domesticated, and the other parts are foreignized.To develop a better insight into the procedures' effects on translational intertextuality, a theoretical framework was presented below.

(Translational) Intertextuality
De Beaugrande and Dressler (1981: 11) offer intertextuality as one of the seven principles of textuality, specializing in the agents which correlate the extant text and previous texts and depend on "the participants' knowledge of other texts " (1981: 182).It contains implicit or explicit intertextual and inter-or intra-disciplinary copulae such as references or quotations to participate actively in the continuity of knowledge and to be regarded as a text.
The translating act itself primarily depends on intertextual links based on interlinguistic dimensions, i.e., the source text and the target text.Thus, it can be said that intertextuality is an essential part of the translation act since a diachronic relationship is assumed between the antecedent text and the extant translation (Kantar 2003: 157).Venuti (2009: 158) offers a vast angle to illustrate the peculiarity of translational intertextuality premised upon "three sets of intertextual relations": "those between the foreign text and other texts", "those between the foreign text and the translation" and "those between the translation and other texts".Venuti (2009: 158) also states that the mentioned three sets are not particularized yet linked in "complex, uneven ways".Therefore, it can be understood that the translator should firstly decontextualize a text prior to translating and latterly recontextualize within the target language's intertextual relations if the subject is already conceptualized in the target language since offering a new piece of information to an unfamiliar recipient can disrupt the communication.It is this decontextualization and recontextualization process that allows a piece of translation comes to establish new intertextual ties -particularly with the recipient system of textual conventions -along with its previous network of textual interdependence (Venuti 2009: 162).
Needless to say, intertextual networks require the dissemination of knowledge, which in turn entails the establishment of textual interrelations.Therefore, authors should be wary of the accuracy of the information imported from another text for the sustainability of intertextual networks, for detected misinformation will not be used by other authors and will potentially disrupt the intertextual chain across works and even end it when authors cease to have recourse to the text incorporating the erroneous information.Yet there is another possibility, which is the spread of this flawed piece of information by unwary quoters (Yıldız, 2021a).Collet (2016: 74) points out that the consequences of a misreading act by a translator will be dire since an individual's misreading only affects themselves, yet a translator's misreading "leads innumerable others astray".In this sense, because a translator involves in the creation of "another intratextual context and another network of intertextual and interdiscursive relations" (Venuti 2019: 162), recontextualization in the target language should be carefully conducted since such textual interrelations may pave the way for the dissemination of false knowledge.Thus, establishing proper retrospective and prospective connections at the intertextual and intralingual levels gains prominence.Yıldız3 (2021b) defines the mentioned connection types: "retrospective intertextuality establishes forward connections between a past text and a current one", and "prospective intertextuality establishes backward connections between a current text and a past one".In addition, Yıldız (2021b) elaborates that prospective intertextuality refers to "a current text's potential connection to a future text and its ability to guide and shape" that yet-tobe-created text.Therefore, translational intertextuality can be theoretically claimed to establish textual interrelations between past, present, and potentially future texts.

Method
This research paper is a qualitative study.It intends to illustrate a translational chain characterized by potential disruptions in translational intertextuality by observing cross-textual alterations from an intra-and multi-lingual perspective on the studied translational phenomena.To do so, the paper was built on the analysis of six translations of an inscription of Kay-Qubādh I of Rûm in the Alanya Shipyard, a Seljukian shipyard built in Alanya, Türkiye.
Qualitative research methods aim to understand the phenomenon of human living and surrounding factors, e.g., society, and thanks to the employment of qualitative methods, researchers plan to reach a more profound comprehension of the targeted phenomenon (Birdwood et al. 2015: 179).In this research study, the purposively sampled contents of the mentioned inscription and its translations were analyzed within the historical discourse.
Qualitative content analysis can be defined as an empirical and systematical analysis of written, verbal, and visual documentation within its own discourse (Mayring 2000: 2).It features the content of a document and intratextual conceptualization, such as main ideas.This method, as Mayring (2000: 2) suggests, enables a text to be re-oriented in tandem with the intentions of analysis.Ezzy (2002: 83-84) states that content analysis commences with units' sampling out of a (textual) population, the definition of the units and determined categories, the decontextualization, categorization and re-analyses of the defined units, the recontextualization of the units based on predetermined categories, and the "interpretation of results".

3
The second author's conference paper that was presented at the Łódź-ZHAW Duo Colloquium on Translation and Meaning, Winterthur, Switzerland on September 2, 2021.
Qualitative purposeful sampling, or relevance sampling, is the effective utilization of "information-rich cases" out of finite resources (Patton 2001: 237;Palinkas et al. 2015: 534).It can be stated that this technique acts in congruence with the paper's intentions by targeting limited but information-rich cases (Palinkas et al. 2015: 534).To illustrate, the ad hoc corpus provides knowledge from limited resources and continues to sample until a satisfactory phraseological repertoire has been created to draw inferences concerning the investigated parameters.Therefore, it can be expressed that it is a time-saver and output-oriented technique since it will endorse the aim of an author.Krippendorff (2004: 119), supporting the above descriptions, gives an account of purposive sampling, which involves a researcher's systematic truncation of the available resources likely to be included in a corpus and suggests that the resultant resources should possess the relevant information that can be operationalized for the purpose of the study.
Building on the foregoing methodological discussion, the authors purposively sampled a corpus of seven translations to dissect them into comparable units, which were then categorized into relevant groups to offer a well-organized interpretation of the translational/textual parameters influential in the corpus.

Corpus
The authors created a corpus of six academic works, including six translations of the analyzed inscription, to illustrate the researched translational phenomena's effects on translational intertextuality.The corpus consists of five books and one master's thesis.These works were purposefully selected to include the multilingual translations of the inscription of Kay-Qubādh I of Rûm placed on the northern entrance of the Alanya Shipyard.Eight works were identified to involve the inscription in Arabic; however, two 4 were found not to include a translation.It was observed that six academic works amass five different languages, including the source language.These languages are Arabic, German, English, French, and Turkish.Table 1 shows the works and the translations.

Analysis
The authors comparatively analyzed seven translations of the inscription of Kay-Qubādh I in the Alanya Shipyard, which were isolated from six academic works (Table 1).The primary aim of the authors was to describe the effects of the translational chain, particularly relay translation, on potential disruptions in translational intertextuality through lexical and syntactical omissions, additions, corrections, and substitutions in the translations as per the mentioned theoretical framework.Riefstahl (1931) was considered to be the initial source of the translations of the inscription due to the absence of further elaboration and in view of Riefstahl's (1931) statement that the inscription had not been published until his work in 1931.

Initial Source: Riefstahl's (1931) German and English Translations
Rudolf Meyer Riefstahl, an American of German origin, specializes in Turkish-Islamic Arts History.In his book, Turkish Architecture in Southwestern Anatolia (1931), the first part focuses on Turkish-Islamic architecture in Anatolia by imparting information on their architectural styles, periods, inscriptions on them, and miscellaneous information regarding the constructions.The second part of the book concentrates on the inscriptions 7 Our itemization of the Turkish translation.8 Our itemization of the English translation.
and their translations located on the structures.The epigraphic writing of the inscriptions in the Arabic alphabet was performed by Orhan Şaik Gökyay, a Turkish poet and linguist, historian, and philologist.The inscriptions were translated into German by Paul Wittek, an Austrian historian, specialized in the Ottoman Empire.The German translations then were translated into English by Riefstahl 9 .Thus, the following renditions are comparatively analyzed by considering Riefstahl (1931) to be the primary source.Arslan, the co-regent of the Commander of the Faithful.

RIEFSTAHL (
Table 2. Translations by Riefstahl (1931) The authors conclude that Riefstahl's English rendition is a relayed translation because both German and English translations aim to reach a genuine audience.Therefore, the English translation cannot be labeled as an indirect translation but relayed since German serves as a relay between the anterior (Arabic) and posterior (English) texts, and both have an authentic readership, even though they were published concurrently.Thus, it can be stated that the inscription was first translated to German and then to English.Moreover, having been published concurrently with the German rendition, Riefstahl's (1931) English translation, from a broader perspective, roughly exhibits the same syntactical organization as the German translation, although the languages are different.It can be concluded that Riefstahl's translation, published concurrently with the German translation, does not endure the transitory nature of language, e.g., changes in nonepistemic values and lexical/phrasal choices exercised during the twentieth century.
It is observed that both Riefstahl and Wittek adopted foreignization as a translation strategy to overcome the difficulties posed by the Arabic titles.On the one hand, they translated the fifth segment without applying foreignization; on the other hand, they did 9 This information was collected from the catalog inputs of BOUN Library, Middle East Institute, and Manchester University.
not translate the third segment, in which Kay-Qubādh I of Rûm is promoted as "the Great" 10 in terms of religion and military.The rationale behind translating the fifth but not the third might be their further discussion on the caliphate-related titles used by the Sultans of Rûm.Moreover, it is detected that the Latinization of the Arabic titles is the same in both translations.

Relay Translation and Retranslation: Combe et al.'s (1931-1956) French Translation and Lloyd and Rice's (1958) English Translation
Répertoire Chronologique d'Epigraphie Arabe (RCEA) ) is a collection of Arabic inscriptions, their translation into French, and miscellaneous information on the inscriptions.The publisher of the book is Institut Français d'Archéologie Orientale (French Institute for Oriental Archeology in Cairo).Even though the institute is located in Cairo, Egypt, the collection also features Anatolia and Mesopotamia.Combe et al. (1931Combe et al. ( -1956: 84) : 84) provided the epigraphic writing and French translation of the inscription of Kay-Qubādh I of Rûm in the Alanya Shipyard.The inscription's reproduction and publication are attributed to Riefstahl (1931) on the same page.One of the writers of this collection, David Storm Rice, posteriorly contributed to Alanya ('Alā'iyya), focused on the Turkish-Islamic architecture in Alanya, Türkiye, and this inscription could be found in this book by Rice.Table 3. Translations of Combe et al. (1931Combe et al. ( -1956) ) and Lloyd and Rice (1958) 10 The phrase "Alā'ed-dunjā wa'd-din Abu" connotes Kay-Qubādh I's greatness in terms of religion and political power and can roughly be translated as "the supreme ruler of the Earth and religion" [our translation based on Yardım (2002: 88)].
Firstly, it was detected that the first segment of the inscription was omitted in the French translation although it was shown in the epigraphic writing in Arabic.In addition, it was found in the epigraphic writing that Combe et al. (1931Combe et al. ( -1956) made a morphological addition, a shadda 11 , to the first segment in contrast to Riefstahl's (1931) epigraphic writing of the inscription (Fig. 1).It was found that the absence of the diacritic sign poses no crucial problem since both miss a shadda in the second word, i.e., "Allah" 12 .Moreover, it was observed that the foreignization technique 13 , primarily applied by Riefstahl (1931), was employed in the third and fourth segments of the translations, which yet include differences in writing 14 .Similar to Riefstahl (1931), Combe et al. (1931Combe et al. ( -1956) ) used the domestication technique in the fifth segment, i.e., the juxtaposition of the political power of Kay-Qubādh I of Rûm and the Abbasid Caliphate.Lastly, as the sub-title suggests, Combe et al. (1931Combe et al. ( -1956) is a relayed translation since the translation was produced from a mediatory language (German or English) to another language (French) in addition to German and English.
Figure 1.The first segments in Arabic in Riefstahl (1931) and Combe et al. (1931Combe et al. ( -1956) ) It can be observed that Lloyd and Rice (1958) do not adopt the foreignization technique in the second segment compared to Riefstahl (1931) while Lloyd and Rice's (1958) translation resemble Combe et al.'s (1931Combe et al.'s ( -1956) ) thanks to their use of semantically similar "auguste/august" and "l'associé/partner" and their Latinized rewording of "Shāhinshāh".In addition, it should be highlighted that Riefstahl uses "exalted" rather than "august"; however, Combe et al. (1931Combe et al. ( -1956) ) and Lloyd and Rice (1958) prefer a lexical substitution by offering "august".Furthermore, Combe et al.'s (1931Combe et al.'s ( -1956) ) omission of the first segment illustrates that Lloyd and Rice do not merely depend on the mentioned sources since the first segment is syntactically reorganized compared to Riefstahl (1931).Thanks to the changes in translational strategies, lexical substitutions, syntactical reorganizations, and Latinized rewordings, Lloyd and Rice (1958) can be labeled a retranslation by partially interlarding the translations of Combe et al. (1931Combe et al. ( -1956) ) and Riefstahl (1931) due to the mentioned textual overlaps.Lastly, it should be remembered that D. S. Rice contributes to both RCEA  and Alanya ('Alā'iyya) (1958) since there are notable similarities in terms of Latinization of the Arabic titles and the use of foreignization.

Retranslation and Relay Translation: Lloyd and Rice's (1958) English and Sinemoğlu's (1964) Turkish Translation
Alanya ('Alā'iyya), written by Lloyd and Rice (1958), amasses the Seljukian inscriptions in Alanya, Türkiye.The inscriptions part of the book provides the Arabic texts, their translations and descriptions, and references if any.The Arabic inscription was attributed to Combe et al. (1931Combe et al. ( -1956)), Riefstahl (1931), and Konyalı (1946).Lloyd and Rice (1958: 55) mention Riefstahl's discussion on the caliphate-related titles and the date of the engraved inscription.Alanya ('Alā'iyya) (1958) was translated into Turkish by Nermin Sinemoğlu in 1964, and it was republished in 1989.It was already rationalized based on their translation preferences, discussed in the previous section, that Lloyd and Rice's (1958) English translation of the inscription can be considered a retranslation.Hence, the Turkish translation can be classified as a relayed translation based on a retranslation (English) since Lloyd and Rice (1958) acknowledge the previous sources, and there are textual overlaps with Combe et al. (1931Combe et al. ( -1956)).Arslan müminlerin emîrinin ortağı.
Table 4. Translations of Lloyd and Rice (1958) and Sinemoğlu (1964) Differently from the mentioned translations, Lloyd and Rice (1958) and Sinemoğlu (1964) apply the domestication strategy in the anterior translation (English) by stating "Allah" instead of "God".The posterior translation (Turkish) shows partially similar indications.To exemplify, in the case of the third and fourth segments, both translations have the same long vowel symbols (¯), but the long vowel symbol was omitted in the fifth, although writing with long vowel symbols is uncommon in the Turkish language.Nevertheless, in the case of the third segment, Sinemoğlu (1964) offers "şehinşah" -a Farsi loan word used in Turkish meaning "the king of kings" 15 -instead of representing 15 Our translation of the definition of "şehinşah" (https://www.etimolojiturkce.com/kelime/şehinşah)(Turkish).
In the first segment, it can be observed that a syntactical reorganization changed the meaning of the introduction of the inscription, i.e., a verbal invocation to God.To elaborate, Riefstahl (1931: 101) translates the first segment as follows: "Thanks are due to God!".Nevertheless, Lloyd and Rice (1958), preferring roughly the same syntactical organization, replaces "Thanks" with "Grace" whereas they translated "God" into "Allah".In addition, Azzam (2017) prefers "Favor" in contrast to "Thanks" and "Grace" by citing Combe et al. (1931Combe et al. ( -1956) ) even though this part was omitted in the French translation.This differentiation in the preference of "grace", "thanks" and "favor" presents a religious dilemma, a sociocultural delusion, and semantic shifts for prospective intertextuality.It is further argued in the translational differences part of the present paper.

Relay Translation and Retranslation: Sinemoğlu's (1964) Turkish Translation and Yardım's (2002) Turkish Translation
Alanya Kitabeleri (2002) features the history of Alanya and miscellaneous information on the inscriptions in Alanya.The book focuses on the Ottoman and Seljukian inscriptions in the region by providing their Latinized transcriptions, Turkish translations, and Arabic epigraphic writings.Ali Yardım, the author of the book, is a scholar specialized in Hadith, inscriptions, and manuscripts.Thanks to his understanding of Arabic, Farsi, and French, Yardım has studied and provided information on the inscriptions located in Alanya where he was born.On the studied inscription, Yardım (2002: 88) does not cite any source regarding the Latinized inscription, Turkish translations, and Arabic epigraphic writing.Yardım (2002: 88-89)  Arslan oğlu Keyhüsrev oğlu Keykubâd.
Table 5. Translations of Sinemoğlu (1964) and Yardım (2002) The studied inscription was first translated into Turkish by Sinemoğlu (1964) through an English relay, namely Lloyd and Rice (1958).However, it should be reminded that the Turkish translation of Sinemoğlu (1964) involves details not found in the English translation.Therefore, Sinemoğlu (1964) must have interpreted several details -mentioned in the relevant part of this paper -from the Arabic epigraphic writing.Moreover, Yardım's (2002) translation illustrates that a plethora of semantic shifts occur in the Turkish translation of Sinemoğlu (1964).Yardım's (2002) Turkish translation can be regarded as a retranslation because it cites none of the sources and due to the domestication of several segments of the Arabic inscription for the sake of the reader, e.g., the translation of Arabic titles and lexical additions.
It is concluded that, up until Yardım's (2002) rendition, the phrase of "Alā'ud-dunyā vad-din Abū" 18 has not been translated and it was observed that the authors prefer to use this Latinized version of the title with minor changes.It is noticed that this technique disrupts the communication between the author and the users of English, French and German languages since there are multiple calques or loan words in the mentioned translations when compared to Yardım's (2002) Turkish translation.In addition, it can be stated thanks to the Latinized inscription of Yardım (2002) that the third segment is more comprehensible in Arabic since four loan words -i.e., "Mu'azzam, alâ, dünyâ, dîn" -are used in Turkish.Nevertheless, Sinemoğlu's (1964) Anglicization of the Arabic title in the Turkish translation arouses a foreignization feeling for a native Turkish speaker.
In the fourth segment, a sentential substitution is noticeable in Yardım's (2002) Turkish translation and another foreignization attempt in Sinemoğlu (1964).Six translations were observed to preserve the Arabic syntactical structure in their translations, 16 Our translation of "En büyük sultan; Muazzam şehinşah".17 Our translation of "Yüce sultan, büyük hâkān"."Hâkān", a Turkish word, is a title given to "the king of kings".For more, please see Turkish Language Association Dictionary (TLA) (sozluk.gov.tr) (Turkish).18 For more information, please see the third footnote.
although intratextual cohesion was disregarded due to the verbatim translation.To exemplify, a caliphate-related title, bestowed by the Abbasid Caliph upon Kay-Qubādh I of Rûm, was placed at the end of the inscriptions' translations as in the Arabic source text; however, a reader of the translated languages may correlate the title with a previous lexical unit, i.e., Kılıç Arslan I of Rûm who is the grandfather of Kay-Qubādh I. Nevertheless, Yardım (2002) conducts a sentential substitution and translation of "'Alâ'üddünyâ ve'd-dîn" and "Ebû' l-feth" 19 to establish intratextual consistency.Lastly, it was observed that Sinemoğlu (1964) uses "b." -an Arabic abbreviation for "son of" -in the Turkish translation and prefers the chronological arrangement of the paternal relations from the preceding to the succeeding ruler.Yardım (2002) uses -"oğlu" 20 -the Turkish equivalence of "b." for domestication and provides the mentioned relations starting from the successor to the predecessor.

Relay Translation and Retranslation: Combe et al.'s (1931-1956) French Translation and Azzam's (2017) English Translation
The Seljuks of Anatolia: An epigraphic study, a master's thesis, focuses chiefly on the semantic aspects of the inscriptions through chronologically assessing the reign period of the Sultans of Rûm.The thesis, in addition, aims to illustrate the strength of the respective ruler by analyzing his title(s) and to correlate the political events and the new titles.It was written by Salma Moustafa Azzam. Azzam (2017) provides rudimental information on the inscription and its place in the Alanya Shipyard.As the source of the inscription, Azzam (2017) cites Combe et al. (1931Combe et al. ( -1956)) Arslan, the partner of the Commander of the Faithful.
It is observed that Azzam's (2017) epigraphic writing of the inscription differs from Combe et al.'s (1931Combe et al.'s ( -1956) ) in terms of the placement of the Arabic diacritics.To exemplify, the difference between Riefstahl's (1931) and Combe et al.'s (1931Combe et al.'s ( -1956) ) was already pointed out, yet Azzam (2017) did not add a "shadda" to the first segment despite citing RCEA  as the source of the inscription.Although this missing shadda was observed to lead to no semantic disturbance, the absence or inclusion of Arabic diacritics in an inscription could notably affect the network of translational intertextuality since diacritics may cause lexicological and thus semantic alterations.Azzam (2017) offers a different Latinized inscription for the first segment in contrast to Yardım's (2002Yardım's ( ), i.e., "al-mina l'-illah" (2017: 72): 72).

"Grace", "Thanks", and "Favor"
In the first segment, it was detected that three different words were used in three English translations, namely, Riefstahl (1931), Lloyd andRice (1958), andAzzam (2017).It was observed that the syntactical structures of the translations of Lloyd and Rice (1958) and Azzam (2017) demonstrate substantial textual overlaps in spite of varying lexicological preferences, while Riefstahl (1931), the main source, changes the overall meaning by a lexicological addition.Furthermore, thanks to the Turkish translations of Sinemoğlu (1964) and Yardım (2002), a potentially incorrect word choice was noticed, and a problematic textual directionality was detected between the Turkish translations, Riefstahl's (1931) two translations, and the English translations by Lloyd and Rice (1958) and Azzam (2017).To elaborate, the potential incorrect word preference is the use of "grace" in an Islamic-oriented context despite the word's close association with Christianity.Another issue is concerned with the directionality of lexical reference, in other words whether the "grace/thanks/favor" should be directed from humans to God or from God to humans.
"Grace" as a word is mainly associated with two meanings: short prayers before or after a meal and God's generosity -including salvation and sanctification -upon humanity.Under the nomenclature of Christianity, the word is frequently used in the Old and New Testament initially as " ḥēn" and "charis" (Esser 1976: 117-118).It is observed that the word, in several parts of the New Testament, was used in different senses thanks to the translation of the Greek-based word "charis".To exemplify, it is noticed that the word was used in several senses, including the following two: "the essence of God's decisive saving act in Jesus Christ […]" and power deriving from God or from Christ (Esser 1976: 119).Moreover, due to these translational differences and changes in theological thinking and conventions over time, "grace" has turned out to be a theological controversy between the denominations of Christianity, e.g., "means of grace" argument between Catholicism and Protestantism (Brittanica 2021).
The reason behind Riefstahl's (1931) and Azzam's (2017) choices of "thanks" and "favor" may be the close correlation of "grace" with Christianity, as illustrated above.Presenting a similar phenomenon through a comparative reading, three verses of the Qur'an, in English, are selected to involve or illustrate similarity with "grace" in religious context yet translated to Turkish as the same word: "With the grace of Allah […]" 22 (Verse 58 of Yunus), "It is Allah's bounty […]" 23 (Verse 4 of Al-Jumu'ah), and "Allah doth confer a favour […] 24 " (Verse 17 of Al-Hujurat).It can be observed that "grace", "bounty", and "favour", within their own context, were translated into "lütuf" as per the Qur'an translation of the Presidency of Religious Affairs of Türkiye."Lütuf" -sometimes "Lutuf" in Turkish 25 -is a Kalam-related word regarding God's deeds to help humanity.
It was seen that Sinemoğlu (1964) uses "minnet" in place of "grace".Turkish Language Association draws a strict line between the meanings of "lütuf" and "minnet": the former denoting "help or a benefit coming from a superior or a respected 26 ", the latter referring to "a feeling of onus because of a favor 27 ".Therefore, in Turkish, "minnet"similar to "thanks"-is directed from humans to God while "lütuf" -similar to "grace" and "favor"-originates from God as far as the analyzed translations are concerned.Moreover, "lütuf" highlights God's omnipotence to bestow favors to humankind.To sum up, "thanks" is directed to God from humanity, and "grace" and "favor" merely of God.This difference is observable in the Turkish equivalences of "grace/favor" and "thanks", which are "lütuf" and "minnet", respectively.
humankind.Therefore, the preference for "minnet" so as to establish equivalence with "grace" in Turkish translation may cause a sociocultural delusion since it is intended to mean that God is the origin of "grace" or "favor" according to the English translations of Lloyd andRice (1958), andAzzam (2017).As pointed out, "minnet" is a feeling of thankfulness for a favor; on the other hand, "lütuf" is help or benefit deriving from a superior or a respected/awed being.In this sense, a great many syntactical and lexical changes are needed in the Turkish translation to establish an intertextual consistency with the previous English translations.
Nevertheless, thanks to Yardım's (2002) Latinized transcription of the inscription, it was observed that the Latinized Arabic transcription of the first segment is "el-Minnetü lillâh" 28 .The verbatim translation of the mentioned Latinized Arabic phrase is "Thanks [or favors] are to God" 29 .Hence, to the authors' best knowledge, it can be suggested that Sinemoğlu's (1964) and Yardım's (2002) Turkish translations are acceptable; however, the English translations of Riefstahl (1931), Lloyd andRice (1958), andAzzam (2017) are assumed to form translational intertextuality based on the misinterpretations of "grace", "thanks" and "favor".The misconceptualization is concerned with whether these three concepts are directed from God to humans or from humans to God. Sinemoğlu (1964) and Yardım (2002) disrupt the translational intertextuality by conducting a verbatim translation from Arabic into Turkish rather than from a mediating language.

31
Throughout the study, Azzam's (2017) Latinization of the titles is preferred unless stated otherwise.
To the authors' best knowledge, the "qasim" formula is the highest title given by the Abbasid Caliphate to a sultan of Rûm.It is followed by the "burhan" and then "nasir" formula.Nevertheless, there are speculations on the rationale behind the bestowal of such a grandiloquent title onto Kay-Qubādh I by the Abbasid Caliphate since the exact date of the bestowal cannot be determined, and it is seen that "burhan" and "qasim" titles were attributed to Kay-Qubādh I on two different engravements in the Sultan Han near Aksaray, Turkey (Riefstahl 1931: 95).In addition, Rogers (1998: 376) implicates that on what grounds the sultan used this title has not been resolved yet.Yardım (2002) makes an adjectival addition -"political partner […]" 32 -when compared to Sinemoğlu (1964).It is thought that Yardım's (2002) adjective addition is made purposefully to illustrate Kay-Qubādh I's "political" prominence in relation to the Arabic title and the period's status quo.Nonetheless, TDV Encyclopedia of Islam offers a translation for the "qasim" formula as "the (all-regards) partner of the Abbasid Caliph" 33 .Therefore, Yardım's (2002) adjective addition delimits the title's pivotal function.In addition, Riefstahl's (1931) "mitregent/co-regent" might be delimiting as well since "regent" as a noun is closely associated with political contexts, and it undermines the king's supremacy of the Abbasid Caliphate or Rûm Sultanate.Knight (1845: 620) defines "regent" as "the person who exercises the power of a king without being king, and the office such a person, or the period of time during which he possesses the power".Moreover, it is stated that the preference for a regent is only executed in cases of illness, mental incapacity, minor heir, and absence from the realm (Knight 1845: 620).

34
The head of state in Islamic or Arabic countries (our translation from TLA Dictionary) (sozluk.tdk.gov.tr)(Turkish).
Turkish reader because of the everyday use of the mentioned words in Turkish.It is seen that multiple works in the corpus provide estimated dates for the inscription by hypothesizing about the use of the title "qasim" in the relevant engravements.Riefstahl (1931) states that the inscription might have been engraved between A. H. 36 626 and 634 37 because of the "qasim" formula.Endorsing Riefstahl's (1931) hypothesis by citing him, Lloyd and Rice (1958: 55) substantiate this stance by mentioning that the "burhan" formula was preferred in different engravements dated to A. H. 625-626 38 in the Alanya Shipyard.Therefore, it can be understood that the "qasim" title was bestowed upon Kay-Qubādh I towards the end of his reign 39 .The primary reason behind the presence of multiple dates or a wide temporal range for the title can be accounted for by the political hiatuses.Azzam (2017: 73) gives an account of a military interaction between the Abbasid Caliph and Kay-Qubādh I without specifying the date.In addition, it was observed that several Turkish academic works include that account without providing an exact date 40 .It is speculated that the Abbasid Caliph sent an envoy to request Seljukian soldiers against the threat posed by the Mongols, defeating Khwārezm-Shāh and proceeding towards Baghdad (Azzam 2017: 73).Kay-Qubādh I provided the Abbasid Caliphate with more soldiers than requested, yet the posed threat was lifted thanks to a military intelligence report stating the Mongolians had decided not to attack.The Abbasid Caliph sent another envoy to notice the soldiers that they would return with gifts to the Kay-Qubādh I (Uymaz 2019: 66-67).

6
It is clear that Sinemoğlu (1964), which uses the title's possible dates of use as available in Riefstahl (1931), miswrites the date in the commentary part of the inscription.It was observed that Riefstahl (1931: 101) provided the date as "626-634 A. H." and Lloyd and Rice (1958:55) specified that the inscription could be engraved after " [A. H.] 623" yet it was realized that Sinemoğlu (1964:61) writes "[A.H.] 523 (1226)".In this sense, Yardım (2002) does not acknowledge their possible date mentioning it in the body text and through a footnote.In this footnote, it is stated that "in this last work [Lloyd and Rice,p.61],there is a mistake about the numerals of the date: […] 523 (1226) must be 623" 41 (Yardım 2002: 89).Thus, it can be stated that Yardım (2002) alleges that Lloyd and Rice (1958) have miswritten the date based on the Turkish translation by Sinemoğlu (1964) without directly referring to Lloyd and Rice (1958), where the date "623" occurs on page 55 in contrast to Sinemoğlu (1964:61).This is why Yardım (2002) assumes that Riefstahl (1931) may have miswritten the date since Sinemoğlu (1964), the Turkish translation of Lloyd and Rice (1958), cites Riefstahl (1931).
6.2.7.An Overview of the Intertextual Trajectory Riefstahl (1931) was already claimed to be the primary source of the Arabic text and the first translations from the source language.Riefstahl (1931) translated the inscription from Arabic (L1) to German (L2) and then English (L3); a relayed translation.Combe et al. (1931Combe et al. ( -1956) ) must have translated the inscription from German or English to French (L4); a relayed translation.Lloyd and Rice (1958) were assumed to translate the inscription from Riefstahl's (1931) English or German or Combe et al.'s (1931Combe et al.'s ( -1956) ) French translation to English; a relayed translation or retranslation.Lloyd and Rice (1958) cited both mentioned works.Azzam (2017) was assumed to translate the inscription from its French version to English; retranslation.Azzam (2017) merely cited Combe et al. (1931Combe et al. ( -1956)).Sinemoğlu (1964) is the translation of Lloyd and Rice (1958) into Turkish (L5); an unmediated translation of the book and a relayed translation of the inscription.Yardım (2002), another Turkish source, does not mention any of the sources as regards the inscription's translation; however, it was observed that Yardım's (2002) translation might have been influenced by Sinemoğlu (1964); allegedly a retranslation.

Conclusion
This is the first paper to present the impacts of relay translations and retranslations upon translational intertextuality by observing semantic alterations and vague translational procedures in a corpus of six historical academic works.It was attempted to illustrate translational intertextuality via retranslations and relay translations, and seven translations, collected through purposeful sampling, which were analyzed through qualitative content analysis.The analysis showed that several of the analyzed translations differed in lexicological and syntactical preferences, yet it was observed that retranslations/relayed translations adopted source text-orientedness as a strategy which was used in the previous translations as in the case of "'Alā'ud-dunyā wad-din Abu'l-Fath".Hence, the persistent use of the same strategy reveals the textual interdependence between six translations in the corpus.Three significant differences were observed in multilingual translations, which are likely to disrupt current and prospective intertextuality.The dilemma over "grace/ thanks/favor" as to whether it is directed from humans to God or bestowed upon humanity by God was thoroughly discussed, and its potential reflections on different cultures were mentioned.Secondly, the analysis of the translations attempting to establish equivalence with a caliphate-related Arabic title illustrated that lexicological additions/preferences could delimit the source text's meaning, foreignize the text for the target audience and incorporate inaccurate conclusions from the substituting words.In these two phenomena, it was realized that four translators chose not to create retrospective intertextual ties even in the direct translation of Sinemoğlu (1964), which disrupts the inscription's intertextuality because the lexical preferences vary across the works in the corpus, as in the case of "grace/thanks/favor".The consequence is the use of three different words for an Arabic word -"minnet" -and multiple ways of praise thanks to syntactical alterations.Lastly, an author's allegation that the initial source text incorporates an inaccurate date relying on its translation from a mediatory source, as in the case of Yardım (2002), made the authors think that a misforged ring in the discussed translational chain may misinform the subsequent ones.
It was understood from the analyses that the disruptions in translational intertextuality through the mentioned translational phenomena could mislead a future translation and eventually disseminate false knowledge or cause a sense of unfamiliarity for the target language and culture.This paper makes some unique contributions to the related literature by discussing translational intertextuality by considering three translational phenomena from a historical perspective: translation proper, retranslation, and relay translation.However, future research can investigate retranslations and relay translations' influences on translational intertextuality across different disciplines.

Acknowledgments
We would like to thank Prof. Alptekin Yavaş for sharing with us his priceless knowledge of the political events concerned with the inscription's engravement date(s).
In addition, we would like to acknowledge Kemal Deniz Karabacak's efforts to provide us with several of the academic works in the corpus.
Partner of the Commander of the Faithful.

Figure 2 .
Figure 2. The trajectory of intertextual dependence in corpus
compares his data about the structural details of the engraved inscription and the construction date of the Alanya Shipyard with Sinemoğlu (1964 or 1989) about the structural details of the engraved inscription and the construction date of the Alanya Shipyard.