The Emotional Lexicon of the Qur’an: Preserving Affective Meaning in Human and AI Translations

1. Introduction

Language serves as a powerful medium for expressing human emotions and shaping how people interpret texts (Pennebaker 2011, pp. 3–10) (1). The Qur'an, beyond being the central religious text for Muslims, exemplifies exceptional linguistic and literary artistry  (2). Among its many features, the Qur'an presents an extensive and nuanced vocabulary related to feelings. It addresses recurring themes such as love, mercy, fear, and hope, which are central to both spiritual and moral lessons, using subtle and varied expressions (3) Sometimes specific word choices convey these emotions directly, while at other times rhetorical techniques, vivid descriptions, and contextual hints add further emotional depth  (4).

The language of the Qur’an is characterized by a high degree of complexity and linguistic richness. Qur’anic discourse is not merely a traditional Arabic discourse but rather represents an advanced linguistic level that surpasses everyday Arabic usage in structure, style, and expression (Rahim & Zulkepli, 2025) (5). Intensive use of rhetorical devices such as stylistic deviation, repetition, ellipsis, etc., is considered the main pillar of Qur’anic language. They contribute to shaping meaning and intensifying the semantic effect of Qur’anic discourse, thereby making Qur’anic language a complex communicative system that requires thorough contextual, linguistic, and cultural understanding  (5).  Ahmad and Ghafar (2025) (6) state that Qur’anic discourse analysis depends on a range of stylistic variations, where phonetic, semantic, and syntactic components are combined to produce well -woven discourse. In addition, sentence rearrangement, ellipsis, repetition, and phonetic structure are also used to foster clarity and moral impact on the receiver. They affirm that these stylistic features are not simply aesthetical ones but have a vital function in shaping meaning and promoting thematic coherence within the discourse.  

Translating the Qur’an into other languages often alters how its emotional layers and complex meanings come across. These changes can influence how readers interpret and understand the text. Qur’anic translation presents several difficulties. The task involves more than simply converting Arabic words into those of another language. Translators must also convey the deep connections to Arabic culture, language, and religious context. Many face challenges in reflecting emotional subtleties that are crucial for engaging readers and encouraging spiritual reflection, for instance, the term “rahma/ رَّحْمَةِ ,” which recurs throughout the Qur'an. Although “mercy” is the usual English translation, this word does not fully express the meaning of the original. “Rahma” covers additional concepts, including compassion, kindness, and divine generosity. Classical references such as al-Raghib al-Isfahani’s Mufradat Alfaz al-Qur’an and more recent studies (Abdel Haleem 2005) (2) note these distinctions. In English and other languages, a single word often cannot capture the entire sense. Similarly, translators commonly use “fear” for “khawf/ خوف,” but in the Qur'anic context, it may refer instead to a feeling of deep respect or awe toward God, rather than simple fear (Ibn Manzur 1955, p 9:83; Mir 1989, p 45) (7, 8) .  Shifts in the emotional tone during translation raise further questions about how much accuracy remains and how interpretations may change. The choices translators make regarding language patterns, cultural context, and theological understanding contribute to variation in how emotional aspects appear. Consequently, readers may encounter emotional elements in Qur’anic translations that differ in significant ways from the original text. Moreover, although AI translations are grammatically fluent and stylistically clear, they still tend to fall short in conveying the full emotional, theological, and rhetorical depth achieved in human translations. Therefore, this issue needs to be addressed by incorporating AI-generated translations as a comparative lens alongside established human versions. In doing so, we can examine how effectively each preserves the emotional intensity of Qur’anic language. To this end, this study aims to: (1) Analyse the translation of emotional expressions in selected Qur’anic verses. (2) Evaluate the effectiveness of translation strategies in maintaining theological and emotional resonance. Accordingly, the research questions would be

1. How do shifts in emotional meaning affect the interpretation of the Qur’an in different linguistic and cultural contexts, including AI and human translations? 2. How are emotional terms translated across different versions of the Qur’an, and how do AI translations compare to human translations in preserving emotional and theological nuances? 3.To what extent do AI translations capture the emotional and spiritual depth of the Qur’an compared to human translators?

3. Theoretical Framework

3.1. Hermeneutical Translation Theory (1975; 1992; 1998)

Hermeneutical Translation Theory centres translation on its explanatory dimension. Within this approach, factors shaping meaning, including the translator’s cultural background, language proficiency, and individual understanding of the source, receive special attention. Steiner (1975; 1992; 1998) (9) points out that translation extends beyond a simple linguistic task, requiring explanation and considerable awareness of both context and culture. In Qur’anic translation, focusing solely on word matching is insufficient. Instead, conveying spiritual, emotional, and theological meaning becomes necessary, which makes this theory especially pertinent. This viewpoint allows for the examination of how translators approach the interpretation and conveyance of Qur’anic emotions as they move between different languages and cultural contexts. According to this theory, variations in emotional content across diverse translations often arise from the translator’s personal outlook, religious experience, or understanding of the target audience’s expectations.

Adopting this perspective also assists in considering how changes in emotional meaning may influence interpretation, which aligns with the core of my research question. Steiner’s Hermeneutic Motion includes four related stages in the translation process: trust, aggression, incorporation, and restitution. These do not represent fixed categories but rather fluid phases within the ongoing process of interpretation. “Trust” appears as the translator’s belief in the value and significance of the original text. During the “aggression” phase, the translator questions and sometimes deconstructs the original material. The stage of “incorporation” refers to introducing elements from the original into the new language. Through “restitution,” the translator aims to address any losses and restore equilibrium, showing ethical care for both texts. Together, these phases illustrate how meaning is continuously negotiated between the original and the translation.

3. 2. Functional Equivalence Theory (Eugene Nida’s 1986)(10)

Eugene A. Nida, a pioneering figure in translation studies, introduced the concept of dynamic equivalence in the 1960s as a shift from formal, word-for-word translation toward a model that emphasizes the receptor audience's response. In his early works, notably Toward a Science of Translating (1964) and the Theory and Practice of Translation (1969, co-authored with Charles R. Taber), Nida (20000) defines dynamic equivalence as “the closest natural equivalent to the source-language message,” focusing on the effect of the message on the target audience. The goal was to achieve a response in the receptor similar to that of the original audience.

Nida identified three main aspects of translation: the message’s content, its linguistic form, and the cultural background of the target audience. In this framework, dynamic equivalence is often contrasted with formal equivalence. While formal equivalence aims to keep the structure and wording of the original text, dynamic equivalence looks beyond surface features. During the mid-1980s, Nida’s model underwent a shift in terminology. He replaced “dynamic equivalence” with “functional equivalence.” In From One Language to Another (1986), which he wrote with Jan de Waard, this change was clearly explained. This revision did not simply update the terms. Nida intended to address misunderstandings caused by the word “dynamic.” Some critics took “dynamic” to mean a translation approach that was too loose or creative, overlooking the detailed linguistic and cultural study central to Nida’s method. Even after this change, functional equivalence remained based on the same core ideas as dynamic equivalence. These ideas stress accuracy in both meaning and intended effect. Functional equivalence, however, placed more emphasis on how the message works within the target language and culture.

Nida and de Waard mentioned that functional equivalence does not seek to imitate emotions or dramatic effects. Instead, they argued that the main aim should be the clear and precise communication of meaning, always adapted to the audience's knowledge and experience. Looking at both approaches, dynamic equivalence introduced in (1964 and 1969) marked an early move towards focusing on how readers respond. Functional equivalence (1986) refined this idea with updated terminology, shifting attention from emotion toward how the message functions and aiming to correct past misunderstandings. Rather than favoring word-for-word translation, this theory stresses transmitting both the sense and the intended outcome of the original. It seeks to bring about in the new audience the same emotional and mental responses as those felt by the first readers. In the case of translating the Qur’an, this approach is particularly relevant. It helps scholars judge how well English translations reflect the emotional depth of the Arabic source. By using this perspective, researchers can study how terms like “rahma /mercy” and “khawf /fear/awe” are adapted across languages and cultures. This theory also helps identify where emotional meaning may be reduced or heightened due to differences in grammar, word choice, or cultural outlook. In summary, hermeneutical translation theory helps explain how translators understand and reshape emotional meaning. Functional equivalence theory, in turn, provides a way to assess whether translations maintain the emotional strength of the Qur’an. Used together, these theories offer a strong foundation for my research.

It should be clarified that this study does not approach Nida’s Functional Equivalence Theory as a matter of mere language shift or lexical substitution. Instead, the concern lies in how meaning, function, and emotional impact are conveyed across linguistic and cultural boundaries. Following Nida, the analysis examines whether the translated text can evoke in the receptor audience a response comparable to that of the original audience. In this sense, functional equivalence provides a framework for evaluating not only the lexical rendering of Qur’anic emotional terms but also the preservation of their affective and theological resonance.

4. Method

In this study, researchers used OpenAI’s ChatGPT system, based on the GPT-4 model, to generate multiple English versions of selected Qur’anic verses with artificial intelligence. The study adopts a qualitative approach. Between June and July 2024, the team produced these translations by interacting directly with the ChatGPT interface (OpenAI 2024). Each translation relied solely on zero-shot prompting, as the team did not perform any fine-tuning. Throughout the process, the researchers did not use any external datasets or plugins.

Five Qur’anic verses from the Chapter of Joseph were chosen in this study. The purpose of including AI translations was to provide a comparative perspective alongside established human translations (Yusuf Ali, Pickthall, Sahih International) to assess the degree to which emotionally charged Qur'anic language is preserved or diminished when rendered by a non-human agent. While AI translations demonstrate grammatical fluency and contemporary stylistic clarity, their emotional, theological, and rhetorical depth remains limited compared to human renderings.

The analysis was divided in this way to highlight the distinct contributions of each theoretical framework. Steiner’s Hermeneutic Motion provides a process-oriented perspective, emphasizing the stages of interpretation, trust, aggression, incorporation, and restitution in translation. Nida’s Functional Equivalence Theory, by contrast, focuses on the functional and affective impact of the translated text on the receptor audience. By treating them separately, the study enables the identification of what each framework reveals about the translation of Qur’anic emotional lexicon and then synthesizes their insights in the comparative discussion. This methodological division avoids conflating the two approaches and ensures that their unique explanatory power is preserved before being combined in the final synthesis.

5. Data Analysis

Steiner’s Hermeneutic Translation Theory

The verse “إِنَّ مَعَ الْعُسْرِ يُسْرًا” (Verily, with hardship comes ease) becomes clearer when examined through Steiner’s hermeneutic framework, which moves through four related stages: trust, aggression, incorporation, and restitution. Each stage offers a distinct method of assessing how both human and AI translations engage with the complex emotional and spiritual elements present in the Qur’anic text. In the trust stage, translators begin with the conviction that transferring meaning from the source language into another is both achievable and necessary. This basic trust in the ability to translate central terms, i.e., “al-‘usr /hardship” and “al-yusr /ease”, is evident in all four translations reviewed: those by Yusuf Ali, Pickthall, Sahih International, and the AI-produced version. However, human translators often display a deeper spiritual trust. Their work not only reflects confidence in communicating the message but also a sensitivity to its existential and theological depth. The AI-generated version, on the other hand, demonstrates only a structural or syntactic trust. While this version conveys the literal meaning, it does not engage with the verse's emotional or spiritual layers. Clearer distinctions emerge in the aggression stage, which involves analyzing the source text to uncover its underlying meanings. Yusuf Ali’s translation, “Verily, with every difficulty there is relief,” broadens the interpretation by including “every,” thereby making the message relevant to a wider range of experiences and suggesting an effort to connect with the text more deeply. Pickthall’s translation introduces a sense of movement and continuity through the phrase “goeth ease,” using older English forms to suggest ongoing comfort. Sahih International, in contrast, remains closely aligned with the Arabic syntax, using “will be” to highlight the temporal aspect and emphasize strict adherence to the original wording.

 The AI translation, “Indeed, with hardship comes ease,” adopts a simplified, modern expression that fulfils a communicative function but lacks rhetorical complexity. Each version interprets the text differently, but only the human translations manifest interpretive nuance informed by context and purpose. In the incorporation phase, the translator internalizes the message and reconstructs it in the target language. Here, the differences become particularly striking. Yusuf Ali incorporates a tone of universality and spiritual assurance, using “verily” and “every” to emphasize both certainty and inclusiveness. Pickthall’s choice of classical English by using “Lo! and goeth” reflects an effort to preserve the elevated tone of the Qur’an. Sahih International seeks clarity and linguistic fidelity, using modern standard English while maintaining the verse’s duality. In contrast, the AI translation incorporates only the semantic shell of the message, the basic connection between hardship and ease, but removes the deeper rhetorical, historical, and emotional layers embedded in the Qur’anic context.

Finally, the restitution phase, where the translator offers the text back to the target audience in a new but faithful form, reveals the human translators’ efforts to restore emotional and spiritual resonance.

By incorporating expressions that emphasize inclusion and perseverance, Yusuf Ali presents the verse as a spiritual concept relevant to people across different backgrounds, thereby ensuring broader understanding. Pickthall, employing a polished literary approach, establishes a sense of reverence while retaining the verse’s spiritual character. Sahih International occupies a position between strict word-for-word translation and interpretive fluency, combining careful wording with an interpretation that remains accessible. This method appears to balance accuracy with readability. The AI translation, although consistent with the original wording, does not convey the verse’s intended meaning fully, resulting in a reduced emotional and philosophical effect.

It does not “return” the verse with the same transformative potential as the original, which is the core of Steiner’s concept of restitution. Through this hermeneutic perspective, it becomes evident that AI translation operates predominantly at the surface level (i.e., the linguistic level without interpretative depth), while human translators engage in deeper interpretive work that mirrors Steiner’s philosophical stages. The ability to embody the worldview, emotional force, and theological depth of the Qur’an remains a profoundly human task, one that artificial intelligence, in its current form, cannot reproduce.

Nida’s Functional Equivalence Theory

Nida’s Functional Equivalence Theory provides a valuable framework for assessing how different translations of the Qur’anic verse “إِنَّ مَعَ الْعُسْرِ يُسْرًا /Indeed, with hardship comes ease” succeed, or fail, in conveying not only semantic content but also the intended emotional and spiritual effect. Central to this model is the idea that a good translation should evoke in the target audience a response functionally equivalent to that experienced by the original audience.

This verse, in its original context, seeks to provide comfort and hope to individuals undergoing hardship. The early Muslim community in Mecca, which experienced persecution, regarded this message as especially meaningful. The verse does more than state a fact; it has a direct effect on its listeners, offering spiritual and psychological comfort understood as coming from a divine source. Yusuf Ali’s translation, “Verily, with every difficulty there is relief”, strengthens this emotional aspect through its emphatic and broad language. Including “every,” though not present in the Arabic, appears to extend the promise, implying that relief is always available, regardless of the hardship. This approach corresponds to Nida’s dynamic equivalence, in which the preservation of emotional impact takes precedence over strict literalism. Such stylistic choices can be justified if they help readers feel the intended reassurance.

The formal tone introduced by “Verily,” together with the expanded promise, enhances the verse’s soothing effect and helps readers not only understand but also feel its message. Pickthall’s version, “Lo! with hardship goeth ease”, adopts a literary style that likely matched the expectations of readers in the early twentieth century. However, for readers in later periods, the use of archaic language and rhythm may feel distant, diminishing the verse’s emotional power. While this poetic and scriptural tone marks the translation as sacred or formal, it may also diminish its ability to connect with contemporary audiences.

From Nida’s theoretical perspective, this translation appears to move toward formal equivalence by prioritizing fidelity to the source structure, possibly at the expense of emotional resonance. Sahih International’s translation, “Indeed, with hardship [will be] ease”, shows a more measured approach. This version retains the original grammatical structure and provides clarity about the sequence of events, using brackets to add interpretive detail. By doing so, it subtly conveys that relief both accompanies and follows hardship. While such explicit clarification helps readers understand the intended meaning, it can slightly diminish the immediate reassurance the original wording provides. The AI-generated version, “Indeed, with hardship comes ease”, achieves clarity and straightforwardness. The literal meaning of the verse remains, but the emotional and spiritual depth is noticeably reduced. Its plain style, although easy to understand, does not carry the expressive qualities found in the Qur’anic text. According to Nida’s framework, this rendering appears to favour naturalness in present-day English, but in doing so, it may not achieve the same emotional or spiritual effect intended by the original. Without interpretive framing, the AI output presents the message clearly but fails to create a transformative reading experience. One key difference between translations created by humans and those by AI is the use of cultural sensitivity.

Yusuf Ali and Pickthall, through their distinct stylistic approaches, adapt the verse to their own linguistic and cultural contexts. Yusuf Ali’s translation is marked by clear and inclusive devotional language, while Pickthall adopts formal elements drawn from sacred literary traditions. Sahih International, adopting a more neutral style, provides a version that contemporary Muslim readers may find accessible. On the other hand, the AI-generated translation relies on algorithmic processes, selecting word choices based on probability rather than deeper cultural or spiritual understanding. The result is a translation that is grammatically correct but lacks cultural or spiritual nuance. Nida’s focus on the reader's response further highlights the limitations of AI translation. Human translators can evoke reassurance, affirmation, or even spiritual uplift, whereas AI-generated text may seem impersonal and emotionally muted. While information is presented, the affective aspect and the ability to move the reader tends not to be realized, which falls short of dynamic equivalence. Taken together, these considerations suggest that, while AI models such as GPT-4 can produce clear and technically sound translations, they are unable to capture the emotional, cultural, and theological depth that is required for sacred texts. Human translators, using judgment and cultural understanding, fulfil Nida’s aim for translation that appeals not just to the mind, but also to the heart.

Artificial Intelligence and the Limits of Qur’anic Emotion Translation

The discussion above has already highlighted the limitations of AI translation, which should be reiterated for clarity. Large language models such as GPT-4 can produce translations with correct grammar and structure. However, these systems lack cultural awareness, sensitivity to theological context, and the deliberate approach found in human translators. Therefore, although AI can support translation tasks, it does not replace the interpretive subtlety and emotional depth necessary for conveying the Qur’an’s meaning.

Synthesis and Critical Reflection

The comparative analysis of the Qur’anic verse “إِنَّ مَعَ الْعُسْرِ يُسْرًا” through the lenses of Steiner’s Hermeneutic Model and Nida’s Functional Equivalence Theory reveals the complex demands of translating emotionally and spiritually charged texts. Human translators, informed by cultural knowledge, rhetorical intuition, and interpretive awareness, demonstrate an ability to navigate both the spiritual resonance and the functional purpose of the original text. Their renderings engage with the source message not just as linguistic data, but as a lived emotional and theological reality, thereby achieving a translation that speaks to both the heart and the intellect.

By contrast, AI-generated translation, while syntactically sound and stylistically modern, operates mainly at the surface level of form. It lacks the interpretive agency needed to fulfil the dynamic and hermeneutic expectations set forth by Nida and Steiner. The model’s reliance on statistical probability, absence of cultural awareness, and inability to contextualize spiritual nuance result in translations that, although accurate in denotation, fall short in connotative and affective power.

This analysis suggests that sacred texts, particularly those as emotionally and spiritually dense as the Qur’an, require human mediation that artificial intelligence, in its current form, cannot replicate. The task of translating such texts is not simply linguistic but profoundly ethical, emotional, and interpretive, demanding a depth of understanding grounded in both faith and scholarship.

These reflections align with broader scholarship on the limitations of machine translation and the challenges of rendering sacred texts. Studies in translation technology note that while neural models achieve high levels of syntactic accuracy, they struggle with cultural nuance, literary resonance, and affective meaning (11,12). Qur’anic translation scholarship likewise stresses that the Qur’an demands not only linguistic transfer but also spiritual and interpretive engagement ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++-------------++++-------------------------------+++++(14). This underscores the necessity of human mediation in contexts where theological and emotional depth are indispensable.

Steiner’s Hermeneutic Translation Theory

Steiner’s hermeneutic motion offers a compelling lens through which to analyse the translation of this profound verse, which encapsulates the essence of the Prophet Muhammad’s mission as a universal embodiment of divine mercy. The four interpretive phases: trust, aggression, incorporation, and restitution help clarify how human and AI translations respond to the emotional and theological weight of the verse.

In the trust phase, all four translators and the AI approach the verse with the assumption that the core concepts “raḥmah /mercy” and “al-‘ālamīn /the worlds” can be rendered intelligibly into English. However, the specific lexical choices reveal differing levels of interpretive depth. Yusuf Ali’s translation, “a Mercy for all creatures”, reflects a strong spiritual trust in the verse’s theological content. His choice of “creatures” broadens the scope beyond humanity to encompass all living beings, emphasizing the universal compassion of the Prophet’s mission. Pickthall, in contrast, selects “the peoples,” a more anthropocentric term that implicitly limits the message to human societies. This reflects a more culturally grounded trust, one that centers human communities as the recipients of divine mercy. Sahih International renders “al-‘ālamīn” as “the worlds,” maintaining close alignment with the Arabic plural and preserving the verse’s cosmic scale, while “mercy” is left unembellished, suggesting a high degree of lexical fidelity. The AI translation renders “all beings” as “al-‘ālamīn,” which, while accurate and inclusive, lacks the spiritual gravitas and rhetorical elevation that human translators seek to preserve. The term “beings” is semantically broad but emotionally neutral, indicating that the AI model treats the verse primarily as a linguistic unit rather than a revelatory utterance. Thus, while all versions demonstrate trust in the verse's basic translatability, only the human translations display an interpretive trust aimed at maintaining the original's doctrinal and emotional force.

In the aggression phase, the moment of interpretive analysis, each translator reveals a distinct method of unpacking the theological and rhetorical complexity of the phrase “raḥmatan lil-ʿālamīn.” Yusuf Ali’s translation, “a Mercy for all creatures,” reflects a deliberate choice to universalize the scope of the Prophet’s mercy. His use of “creatures” instead of a more precise term like “people” or “nations” suggests an expansive vision grounded in Islamic theological interpretations, particularly those that view the Prophet’s message as relevant not just to humanity, but to all forms of life, including animals, jinn, and possibly even inanimate creation. This decision demonstrates a cosmological reading, often associated with classical tafsīr, and shows a high degree of interpretive engagement. Pickthall’s choice of “the peoples” indicates a narrower, anthropocentric focus. He avoids more abstract or metaphysical translations, instead grounding the message in human plurality. This aligns with Pickthall’s general preference for clarity and his linguistic conservatism. The phrase “the peoples” suggests discrete human communities, which may reflect a modernist or socio-political understanding of Islam’s universal relevance to different nations and cultures without invoking the more mystical dimensions implied by “worlds” or “creatures.” Sahih International selects “the worlds,” a direct and semantically faithful translation of “al-ʿālamīn.” This choice preserves the original Arabic plural and maintains the theological ambiguity inherent in the verse. “Worlds” is intentionally open-ended—it can refer to the physical cosmos, multiple planes of existence (including the seen and unseen), or different communities of beings.

This lexical choice corresponds with Sahih International’s overall strategy of literal accuracy with minimal interpretive overlay, allowing readers to engage with the text's polyvalence themselves. The AI translation, “all beings,” is grammatically and semantically adequate, offering a broad and inclusive equivalent for “al-ʿālamīn.” However, this lexical choice is drawn from frequency-based pattern recognition rather than theological or exegetical reasoning. While “beings” captures plurality and inclusivity, it lacks the metaphysical depth and historical resonance that guide human translations. The AI model does not recognize that “al-ʿālamīn” is often interpreted in Islamic scholarship to include human and non-human realms, both seen and unseen; thus its rendering, although correct in a superficial sense, lacks exegetical intentionality.

In the incorporation phase, each translator reshapes the Qur’anic message within the linguistic and rhetorical framework of the target language, guided by their interpretive philosophy and intended readership. Yusuf Ali’s rendering, ‘We sent thee not, but as a Mercy…’, demonstrates a deliberate use of elevated, quasi-biblical diction, characterized by the archaic pronoun ‘thee’ and inverted syntax. This stylistic choice reflects his broader strategy of conveying not only meaning but also spiritual gravitas and sacred tone, aligning the translation with English scriptural traditions. Pickthall’s phrase “save as a mercy” parallels this approach; his use of “save as” rather than “except” evokes solemnity and textual reverence, reinforcing the verse’s divine authority while maintaining a more restrained poetic style. Sahih International, in contrast, prioritizes clarity and syntactic fidelity, replicating Qur’anic structure with the phrase “We have not sent you… except as a mercy.” This version ensures formal accuracy and wide accessibility, especially for modern readers and non-specialist audiences, though at the expense of some rhetorical flourish.

The AI-generated version approaches the verse differently, aiming for smooth contemporary English: ‘And We have not sent you except as a mercy to all beings.’ While grammatically correct and fluent, this simplicity contributes to stylistic flatness. The absence of rhetorical elevation diminishes the verse’s performative power and liturgical resonance. Furthermore, the phrase ‘all beings,’ though inclusive, lacks the doctrinal specificity and cosmological breadth traditionally associated with “al-ʿālamīn” in Islamic exegetical traditions. As a result, the AI’s incorporation of the verse emphasizes linguistic naturalness over theological density, delivering an accessible but spiritually neutral reconstruction of a highly charged religious statement.

In the restitution phase, the focus shifts to how effectively each translator returns the verse to the target audience in a form that preserves its original spiritual power, emotional resonance, and rhetorical gravity. The human translators show a clear intention not merely to inform but to elevate, ensuring that the verse retains its performative role as a declaration of divine mercy and prophetic purpose. Yusuf Ali’s choice of “creatures” in “a Mercy for all creatures” reflects an expansive theological vision, aiming to encompass all forms of life within the sphere of divine compassion. This word choice supports his broader goal of presenting the Qur’an in universalist, almost cosmic terms, and helps restore the inclusive majesty embedded in the concept of “al-ʿālamīn.” His use of capitalized “Mercy” and poetic rhythm further reinforces the spiritual gravitas of the original verse.

Pickthall restitutes the message through the phrase “the peoples,” which, though narrower in scope, effectively emphasizes the plurality of human societies and the Prophet’s role in addressing diverse nations. His word choice highlights the social and historical dimensions of the Prophet’s mission while maintaining a reverent tone through his conservative syntax. Sahih International’s choice of “the worlds” stays lexically closest to the Arabic “al-ʿālamīn,” enabling a restitution that reflects the Qur’an’s multi-layered worldview without imposing interpretive detail. By preserving the original's ambiguity, Sahih’s version invites readers to consider both the material and metaphysical scope of prophetic mercy. The AI-generated translation, “a mercy to all beings”, while semantically correct, delivers the verse in a flattened rhetorical register. The phrase “all beings” is inclusive but theologically underdeveloped, lacking the exegetical sensitivity or historical embeddedness found in human translations. Furthermore, the overall sentence structure is stylistically neutral and devoid of the emotive or reverential tone that the Qur’anic context demands. The AI model, operating without interpretive consciousness, restitutes the text as an informational statement rather than a sacred utterance, thus missing the transformative function central to Steiner’s hermeneutic ideal. While the content is technically intact, the affective and theological force is significantly diminished.

This example further illustrates the consistent pattern emerging throughout the analysis: while AI-generated translation achieves grammatical correctness and basic semantic fidelity, it falls short in restoring the spiritual, rhetorical, and theological power of the original. The human translators, through varying degrees of poetic elevation, theological intentionality, and sensitivity to classical exegetical traditions, demonstrate the kind of hermeneutic engagement that sacred discourse requires. Their lexical choices, such as Yusuf Ali’s expansive “creatures,” Pickthall’s pluralistic “peoples,” and Sahih International’s ambiguous yet faithful “worlds,” reflect a deliberate effort to preserve both the scope and sanctity of the Prophet’s universal mission. In contrast, the AI translation, though linguistically smooth, treats the verse as an informational unit, lacking the interpretive understanding required to restitute the Qur’anic message with its original emotive and metaphysical richness. The multi-layered labour of trust, interpretive penetration, cultural incorporation, and spiritual restitution remains firmly within the realm of human translation.

Nida’s Functional Equivalence Theory

Nida’s Functional Equivalence Theory emphasizes the necessity of producing the same effect on the target audience as the original text had on its initial recipients. This requires not only linguistic precision but also cultural and theological adaptation.

In Yusuf Ali’s translation, “We sent thee not, but as a Mercy for all creatures”, the use of archaic pronouns like “thee” and the poetic phrasing “but as a Mercy” reflects a deliberate alignment with the biblical English style familiar to readers of sacred Christian texts. This stylistic choice embeds the Qur’anic message within a sacralized linguistic tradition, enhancing its liturgical resonance and appealing to a spiritually inclined audience. By selecting “creatures” rather than “people” or “beings,” Yusuf Ali emphasizes the universal scope of the Prophet’s mercy, extending it to all living entities, not solely humans, an interpretive decision influenced by classical Islamic theology and tafsīr. Sahih International adopts a more neutral and accessible tone with the phrase “mercy to the worlds,” preserving the theological significance of “al-ʿālamīn” without elaboration. The term “worlds” retains the cosmic breadth of the original Arabic and leaves room for interpretation—whether referring to humanity, jinn, angels, or multiple realms of existence. This lexical choice reflects a balance between doctrinal precision and stylistic restraint, allowing the translation to resonate with a broad contemporary audience while preserving Qur’anic ambiguity.

The AI-generated version, “a mercy to all beings”, prioritizes semantic clarity and modern phrasing, but in doing so, it removes the verse from the linguistic traditions of sacred discourse. The term “beings” is inclusive and accurate in a general sense but lacks the spiritual and eschatological implications that “creatures” or “worlds” carry. Moreover, the sentence structure is flat and informational, reducing the verse to a neutral statement of fact rather than maintaining its performative role as divine proclamation. This reflects the AI model’s tendency to treat text as linguistic data, rather than as revelation intended to evoke emotional and theological reflection.

From a receptor-response perspective, readers of Yusuf Ali’s translation are likely to perceive the verse as a reverent affirmation of the Prophet’s transcendent mission, reinforced by its poetic form and elevated diction. Readers of Sahih International encounter a more literal rendering that nevertheless preserves theological integrity. By contrast, the AI-generated translation, while intelligible and fluent, lacks the devotional tone and rhetorical gravitas needed to convey the compassionate and universal scope of the Prophet’s mission. Consistent with findings from earlier verses, the AI version achieves only surface-level equivalence; in Nida’s terms, it falls short of functional equivalence, as it fails to reproduce the verse’s intended emotional, cultural, and spiritual impact.

Technical Commentary on AI Translation

The AI-generated rendering of “لِلْعَالَمِينَ وَمَا أَرْسَلْنَاكَ إِلَّا رَحْمَةً  ” as “And We have not sent you except as a mercy to all beings” demonstrates the broader limitations of GPT-4 in translating sacred texts. Although grammatically accurate, it weakens the theological nuance of “al-ʿālamīn,” which in Qur’anic usage refers specifically to humans, jinn, and angels, as noted in classical tafsir (e.g., Ibn Kathir n.d.; al-Sa‘di n.d.). Unlike human translators who draw upon exegetical tradition, AI lacks contextual and spiritual awareness. This illustrates that while AI translations may be informative, they fall short in conveying the emotional and theological depth central to the Qur’an.

Synthesis and Reflection

The analysis of Qur’an (107) through the lenses of hermeneutic and functionalist theories highlights the complexity of translating spiritually charged texts. Steiner’s model reveals that human translators engage with this verse not merely as a linguistic task but as a profound interpretive act, internalizing and restituting its spiritual force through choices in tone, scope, and reverence. Nida’s framework demonstrates how emotional and theological impact is preserved or diminished depending on the translator’s sensitivity to the verse's functional purpose, namely, to affirm the Prophet’s universal role as a mercy across all domains of existence. While AI translation succeeds at producing grammatically coherent output, it lacks the interpretive awareness and spiritual intentionality required to fully capture the verse’s doctrinal richness and devotional tone. This case further demonstrates that human agency remains indispensable in the translation of sacred texts, where meaning is inextricably tied to context, theology, and affect.

Steiner’s Hermeneutic Translation Theory

The Qur’anic verse “وَقَذَفَ فِي قُلُوبِهِمُ الرُّعْبَ” (Al-Ahzab 33:26) presents a stark emotional shift from earlier examples, expressing divine-induced fear among the enemies of the Prophet. The verse’s brevity and intensity pose a unique challenge to translators who must convey not only the semantic content but also the forceful emotional impact of the expression. Steiner’s hermeneutic model, encompassing trust, aggression, incorporation, and restitution, provides a critical framework to evaluate how this verse has been rendered into English by human and AI translators.

In the trust phase, all translators accept that the powerful term “الرُّعْبَ”, which denotes a deep, overwhelming kind of terror, can be translated into English. However, the nature of that trust differs. Yusuf Ali and Sahih International both opt for “terror”, suggesting a willingness to preserve the verse’s intensity and moral clarity. Pickthall’s choice of “panic” reflects a slightly different interpretive trust — favouring psychological immediacy over theological force. The AI model, conversely, uses “fear”, which is semantically related but less intense, indicating a minimalist level of trust in conveying the full emotional charge of the original.

The aggression phase, the translator’s intellectual penetration of the source, is apparent in the nuanced choices for “الرُّعْبَ.” Yusuf Ali and Sahih International preserve the Quranic tone by using “terror”, maintaining a sense of divine power and strategic justice. Pickthall’s “panic” introduces an element of disorder and loss of control, reflecting an emotional response but arguably downplaying the verse’s theological framing of divine intervention. The AI's substitution of “fear” suggests a more neutral emotional response, diminishing the verse’s intensity. Each choice reflects a different interpretive aggression — the willingness to engage with or soften the force of the original. Classical commentaries, such as those of Ibn Kathir, highlight this verse as an expression of divine justice and military consequences, a theme evident in the emotional tone of human translations. (Tafsir Ibn Kathir n.d.)

During the incorporation phase, the translator internalizes the meaning and renders it within the target linguistic and cultural context. Yusuf Ali’s construction, “cast terror into their hearts”, adopts an elevated tone fitting for sacred language. Sahih International closely follows this formula, reinforcing both clarity and rhetorical force. Pickthall’s version, “cast panic into their hearts,” reflects his literary sensibilities but reduces the moral weight of the act — presenting it more as an emotional state than a divine decree. The AI translation, “cast fear into their hearts,” lacks stylistic elevation or spiritual gravity. It is functionally accurate but stylistically flat, incorporating only the literal meaning while failing to absorb the Qur’anic tone of judgment and retribution.

In the restitution phase, the translator returns the message to the target audience, ideally with an intensity equivalent to the original. Yusuf Ali and Sahih International successfully restore the original intensity, presenting divine justice in a manner that reflects both emotional gravity and spiritual authority. Pickthall, though literarily expressive, restitutes the verse with a softer psychological register. The AI translation does not achieve full restitution; by choosing “fear”, it dilutes the moral-emotional function of the verse and offers a generalized reaction rather than a specific, divinely instilled terror. The theological depth and emotional resonance of the original are thus only partially restored in the AI’s rendition.

This example confirms the limitations observed in earlier cases: while AI can reproduce the general semantic structure, it fails to capture the emotional and theological function of Qur’anic expressions. The spiritual rhetoric of “الرُّعْبَ”, rooted in divine power and justice, demands interpretive engagement that remains distinctly human.

Nida’s Functional Equivalence Theory

According to Nida’s Functional Equivalence Theory, a successful translation must convey the same effect on the target audience as the original message has on its original recipients. In this verse, “ وَقَذَفَ فِي قُلُوبِهِمُ الرُّعْبَ”, the Arabic evokes divinely instilled terror—not mere fear, but a sudden, overwhelming psychological state that contributed to the collapse of enemy resistance. Thus, any equivalent English rendering must preserve both the emotional intensity and the theological implications of divine action.

Yusuf Ali’s translation, “And cast terror into their hearts,” effectively captures the severity and suddenness of the divine act. The term “terror” evokes a strong emotional response and corresponds well with the verse’s dramatic context. Moreover, his use of “cast”—a verb conveying forceful insertion—adds to a heightened sense of divine power and psychological impact. This rendering fulfils Nida’s criteria for functional equivalence by reproducing the rhetorical intensity and the intended emotional reaction: dread, helplessness, and divine judgment.

Sahih International’s version, “and cast terror into their hearts,” closely mirrors Yusuf Ali’s and is equally effective. Its strength lies in its linguistic precision and emotional clarity, striking a balance between literal fidelity and impactful delivery. From a functionalist perspective, this translation succeeds in replicating the shock and emotional force of the Arabic in clear, accessible English, especially for contemporary audiences.

Pickthall’s rendering, “and cast panic into their hearts,” introduces a subtle shift in tone. The word “panic” suggests a disorganized emotional reaction, which is somewhat consistent with the scene described in the Qur’an. However, “panic” carries a slightly informal or even trivial tone in modern usage, which may diminish the spiritual gravity of the verse. From Nida’s viewpoint, the translation only partially achieves functional equivalence. While it conveys fear, it fails to match the reverent tone and theological weight of the original.

The AI-generated version, “And He cast fear into their hearts,” is grammatically correct and broadly comprehensible but lacks the intensity and specificity of “terror.” The word “fear” is generic and emotionally softer than “terror,” potentially reducing the verse’s rhetorical impact. Moreover, the use of “He” rather than “We” (as in many Qur’anic translations reflecting divine speech) introduces a subtle theological ambiguity, distancing the reader from the Qur’an’s own voice. From a functionalist perspective, this version does not achieve the equivalent emotional effect. It communicates divine action but lacks the urgency, magnitude, and fearsome authority intended by the original.

Under Nida’s Functional Equivalence Theory, the goal of translation is not merely to replicate the lexical content of a source text but to recreate its emotional, cultural, and theological effect for the target audience. In this verse, the emotional force lies in God’s act of instilling “ruʿb” (a term connoting intense fear or terror) in the hearts of the enemy. The human translators demonstrate clear cultural and theological sensitivity in their renderings. Yusuf Ali uses “terror,” a word that strongly conveys divine retribution and evokes the gravity of the moment. This choice aligns with his broader strategy of preserving the moral and eschatological weight of Qur’anic discourse. Sahih International also chooses “terror,” maintaining both the semantic precision and rhetorical charge of the source term. Pickthall’s use of “panic” shifts slightly in tone; while accurate, it introduces a more psychological and less theological interpretation of “ruʿb,” suggesting confusion or disarray rather than the awe-inspiring fear associated with divine intervention. The AI translation, “fear,” is semantically correct but emotionally and spiritually diluted. The term is too general to capture the intensity and divine intentionality embedded in the Arabic, and it reflects the model’s tendency toward lexical safety over emotional specificity. Lacking theological context or interpretive depth, the AI rendering fails to recreate the psychological and moral impact that the verse carries in its original context. In terms of receptor response, human readers encountering Yusuf Ali or Sahih International are more likely to perceive the verse as a manifestation of divine power and an act of moral reckoning, particularly in the context of prophetic warfare and divine justice. Yusuf Ali’s use of “terror” does not merely inform—it arrests the reader and evokes fear, aligning with the Qur’an’s dramatic rhetorical function in recounting historical judgment. Sahih International achieves a similar effect through its lexical fidelity and formal tone. In contrast, the AI version, while grammatically smooth, presents the verse as a neutral event, lacking emotional urgency or sacred tone. The term “fear,” though correct on a basic level, does not provoke the same visceral or spiritual response and reduces a moment of divine confrontation to a mere psychological state. As with previous examples, the AI succeeds in delivering basic denotation, but it fails to achieve dynamic equivalence—the communicative goal of producing an equivalent emotional and spiritual effect on the reader. This highlights a broader limitation of algorithmic translation applied to texts imbued with theological intensity and moral consequence.

Technical Commentary on AI Translation

The AI-generated version, “And He cast fear into their hearts,” reflects the model’s tendency to choose linguistically safe and semantically broad equivalents. While grammatically correct, the selection of “fear” reduces the verse’s emotional and rhetorical intensity, particularly compared to “terror” in human translations. GPT-4, lacking interpretive intent or theological understanding, selects words based on statistical frequency rather than contextual significance. Moreover, its use of “He” instead of “We” diverges from the Qur’an’s divine voice, subtly diminishing the sense of direct revelation and divine authority. The result is a translation that is accurate in surface meaning, yet insufficiently charged with the spiritual force the verse demands.

Synthesis and Reflection

This example further illustrates the central limitation of AI-generated translation: while it may produce fluent and semantically accurate renderings, it lacks the interpretive depth and emotional nuance necessary to convey divinely-instilled dread in a sacred context. Human translators like Yusuf Ali and Sahih International succeed in preserving both the rhetorical intensity and theological weight of the verse, enabling a spiritually resonant experience for the reader. AI, by contrast, operates within a framework of linguistic neutrality, producing a translation that is functionally clear but spiritually diminished. This underscores the continued need for human interpretive agency in translating texts where meaning is inseparable from emotion, context, and revelation.

Steiner’s Hermeneutic Translation Theory

The verse “يَا أَيُّهَا الَّذِينَ آمَنُوا اسْتَعِينُوا بِالصَّبْرِ وَالصَّلَاةِ ۚ إِنَّ اللَّهَ مَعَ الصَّابِرِينَ” functions both as a directive and a consolation, instructing believers to seek divine support through inner endurance and spiritual devotion. Steiner’s four stages — trust, aggression, incorporation, and restitution — provide a lens for evaluating the interpretive quality of each translation.

In the trust phase, all translators demonstrate confidence in rendering the terms “ṣabr” and “ṣalāh” into English. However, the nuances differ. Yusuf Ali uses the phrase “patient perseverance,” a composite expression that attempts to capture both the spiritual endurance and emotional restraint implied by “ṣabr.” Pickthall’s “steadfastness” leans toward moral and psychological resolve. Sahih International and the AI model use the simpler “patience,” reflecting lexical fidelity but less semantic depth. The variation here reflects different degrees of trust not only in the target language’s lexical range, but in its ability to preserve the Qur’an’s rhetorical force.

In the aggression phase (Steiner’s first movement of the Hermeneutic Motion, i.e., the translator’s initial act of engaging and breaking open the source text¹), each translator breaks open the emotional and spiritual layers of “ṣabr” and “ṣalāh.” Yusuf Ali's phrase “patient perseverance” expands the term interpretively, making its spiritual and psychological weight more visible. Pickthall’s use of “steadfastness” aligns with classical English spiritual vocabulary, though it may underplay the emotive burden of enduring hardship. Sahih International and the AI translation rely on “patience,” which is accurate but semantically flatter — it lacks the active connotation of resilience present in the original Arabic. The imperative “اسْتَعِينُوا” is rendered consistently in all versions as “seek help,” showing clear interpretive coherence.

In the incorporation phase, the translations differ in tone and rhythm. Yusuf Ali’s “O ye who believe!” and “patient perseverance and prayer” are clearly influenced by biblical English, adding solemnity and rhythm that match the Qur’an’s spiritual register. Pickthall maintains this formal tone with “Lo! Allah is with the steadfast,” echoing classical religious diction. Sahih International and the AI version use modern, formal English, prioritizing clarity and accessibility. However, in doing so, they lose some of the sacred cadence and elevated register that help carry the emotional authority of the verse.

In the final restitution phase, the translator delivers the verse back to the reader with emotional and spiritual resonance. Yusuf Ali and Pickthall succeed in maintaining a tone of divine reassurance, especially in the final clause: “for God is with those who patiently persevere” and “Lo! Allah is with the steadfast.” Both echo the Qur’anic intent to comfort and inspire the believer. Sahih International’s “Indeed, Allah is with the patient” is semantically sound but lacks rhetorical elevation. The AI version is identical to Sahih’s - grammatically correct, but functionally neutral. It returns the meaning, but not the emotional charge or sacred tone, which are critical in devotional contexts.

In summary, human translators demonstrate deeper hermeneutic engagement with the verse, infusing their renderings with spiritual rhythm, theological sensitivity, and cultural awareness. The AI model remains textually accurate but spiritually and emotionally restrained, reflecting limitations seen in previous examples.

Nida’s Functional Equivalence Theory

From a functionalist perspective, Qur’an (153) is a spiritually motivational directive aimed at inspiring strength, trust, and devotion. Its emotional impact on the original audience, the early Muslim community under pressure and hardship, would have been one of reassurance, encouragement, and theological grounding. The goal of the translation, therefore, is to evoke a similar emotional and spiritual response in the target audience.

Yusuf Ali’s translation, “O ye who believe! seek help with patient perseverance and prayer; for God is with those who patiently persevere”, successfully reproduces this emotional depth. The phrase “patient perseverance” expands on the Arabic “ṣabr,” giving it an active and enduring character. His use of formal, elevated diction (e.g., “O ye”, “for God is with”) aligns the tone with biblical English, reinforcing the spiritual and sacred register. This stylistic choice enhances functional equivalence by reflecting the verse’s role as both command and comfort. While somewhat archaic for contemporary readers, it preserves the solemn authority of the Qur’an.

Pickthall’s version, “Seek help in steadfastness and prayer. Lo! Allah is with the steadfast”, captures the verse’s core emotional force, but with a somewhat narrower lexical rendering of “ṣabr” as “steadfastness.” While this term conveys moral firmness, it is linguistically more neutral and less semantically rich than the broader connotations of emotional endurance and resigned trust implied in “ṣabr.” Nevertheless, Pickthall’s use of formal structures (e.g., “Lo!”) and a reverent tone generates a strong functional impact for readers accustomed to traditional religious discourse. However, this same archaic style may pose accessibility challenges for secular or younger audiences unfamiliar with older English forms, potentially restricting its resonance. A further question arises: does the elevated style and reverent tone compensate for the reduced semantic scope of “steadfastness”? If so, Pickthall’s rendering illustrates how stylistic choices may partially bridge semantic gaps in translation, though perhaps at the cost of accessibility. Sahih International’s translation, “O you who have believed, seek help through patience and prayer. Indeed, Allah is with the patient”, prioritizes semantic clarity and modern linguistic accessibility. The straightforward word “patience” is an accurate, though somewhat restrained, equivalent of “ṣabr.” While it retains the meaning, it does not fully capture the multi-dimensional nature of “ṣabr” in Qur’anic theology, which includes emotional restraint, spiritual endurance, and moral commitment. However, this version offers a clear and dignified rendering, making it functional for readers unfamiliar with Qur’anic depth, though somewhat limited in emotional reach. The AI version, which mirrors Sahih International almost exactly, shares the same strengths and limitations. It is precise, fluent, and formally correct, but lacks any stylistic elevation or emotive enrichment. It does not engage with the verse’s pastoral tone, nor does it attempt to interpret or enhance the emotional or devotional impact. As with previous cases, the AI rendering is basic in intelligibility, but it fails to evoke the emotional reassurance and spiritual resonance that Nida’s theory demands for a functionally equivalent translation. In terms of receptor response, readers of Yusuf Ali or Pickthall may experience the verse as a spiritual exhortation, one that instils both obedience and consolation. Readers of Sahih International (and the AI version) will understand the message but may not feel the same emotional urgency or sacred tone, which are essential aspects of the Qur’an’s communicative function. Thus, this example confirms the pattern: AI translations meet the informational needs of the reader but fall short of delivering the emotional effect, which is critical when dealing with verses meant to uplift, console, or spiritually guide believers.

Technical Commentary on AI Translation

The AI-rendered translation, “O you who have believed, seek help through patience and prayer. Indeed, Allah is with the patient,” reflects GPT-4’s ability to produce clear, grammatically correct English with a neutral tone. This translation is structurally identical to Sahih International’s, likely because of high-frequency exposure during training. While the output is linguistically accurate, it reflects the AI model’s tendency to prioritize semantic precision over rhetorical force. The emotionally rich term “ṣabr” is translated as “patience” without taking into account its spiritual and psychological dimensions, such as endurance, struggle, or sacred resignation. The model lacks interpretive sensitivity, treating the verse as informational instruction rather than divine exhortation, thereby diminishing its pastoral and theological significance.

Synthesis and Reflection

This verse reinforces the broader pattern identified in earlier examples: AI translations achieve surface-level accuracy but lack emotional and spiritual depth. While the GPT-generated version is clear and accessible, it fails to capture the elevated tone and complex force of the Qur’anic appeal to patience and prayer. In contrast, human translators like Yusuf Ali and Pickthall imbue the verse with devotional intensity and reverent language, preserving both the commanding tone and the comforting message. From the perspectives of both Steiner and Nida, AI’s performance underscores the limitations of algorithmic translation in contexts where emotive effect, sacred register, and theological sensitivity are inseparable from meaning.

Steiner’s Hermeneutic Translation Theory

The verse “قَالَ إِنَّمَا أَشْكُو بَثِّي وَحُزْنِي إِلَى اللَّهِ وَأَعْلَمُ مِنَ اللَّهِ مَا لَا تَعْلَمُونَ (86)” is one of the most emotionally significant moments in the Qur’an. It captures the spiritual maturity of Prophet Ya‘qub as he directs his grief exclusively to Allah, reflecting themes of divine trust, private lamentation, and psychological resilience. Steiner’s model provides a useful lens to examine how various translators interpret and convey this delicate emotional register.

In the phase of trust, all translators show confidence in the translatability of the key Qur’anic terms “bath-thī”(generally interpreted as internal sorrow or emotional turmoil) and “ḥuznī” (grief), as well as in communicating the overall message of intimate lamentation directed exclusively to God. However, the nature and extent of that trust differ. Yusuf Ali presents a more interpretive form of trust, rendering “bath-thī” as “distraction,” a choice that leans toward psychological interpretation, suggesting mental fragmentation and emotional fatigue. Pickthall, by contrast, maintains stronger lexical fidelity by translating the terms as “distress and anguish,” which more closely retain the emotional intensity of the original. Sahih International selects “suffering and grief,” a clear and accessible formulation that balances faithfulness and readability. The AI translation (GPT-4) reflects this approach, providing “sorrow and grief” — a safe and semantically precise pair of terms, though arguably less nuanced in emotional complexity and expressive depth.

In the aggression phase, that is, the translator’s level of interpretive intervention (or “translator agency”), the strategies vary considerably. Yusuf Ali’s use of “distraction” introduces a psychological perspective that adds interpretive richness, perhaps intending to modernize or psychologize the emotional struggle for contemporary audiences, yet it risks diverging from the emotional weight carried by the original Arabic. Pickthall adopts a more expressive rendering with the verb “expose,” which evokes a raw and vulnerable emotional gesture, capturing the deeply personal act of spiritual disclosure. Sahih International remains more restrained, using “complain” with minimal interpretive intrusion. This preserves fidelity but yields a more linguistically neutral and less affectively charged tone. The AI translation follows this route, rendering “sorrow” and “grief” with clarity but lacking the visceral emotional texture found in human versions, especially compared to terms like “anguish” or “distress.”

This variation also reflects broader translation philosophies. Pickthall’s choice of “expose” corresponds with a dynamic equivalence orientation, prioritizing expressive force, while Sahih International’s “complain” aligns more closely with formal equivalence. Referencing House’s overt/covert distinction (House 1977, 2015) (15,16) further clarifies that Yusuf Ali and Pickthall tend toward overt, interpretive renderings, whereas Sahih International and the AI remain covert and restrained. Finally, comparing this passage with other translations of “ṣabr” by different translators underscores that the selection of terms such as “steadfastness” or “patience” often balances theological precision with emotional resonance, revealing the translators’ implicit priorities.

Taken together, this emotionally rich verse challenges each translator’s skill to navigate sincerity, theological sensitivity, and emotional depth. The human translations demonstrate a layered hermeneutic process, combining poetic form, affective tone, and interpretive clarity. GPT-4, by contrast, offers a translation that is lexically adequate but emotionally muted, highlighting once more the limitations of algorithmic translation when tasked with conveying the spiritual intensity and existential depth of sacred expression. The verse’s profound expression of grief and divine trust ultimately demands a translator’s empathic insight, not just computational accuracy.

Nida-based Functional Equivalence analysis

According to Nida’s Functional Equivalence Theory, a successful translation should produce in the target audience an effect comparable to that experienced by the original audience. This verse reflects an intensely emotional moment: Prophet Ya‘qub’s grief at the loss of his sons, expressed not as despair, but as spiritual resignation and divine trust. The verse is not merely informative—it is deeply affective, embodying sorrow, resilience, and monotheistic faith in the face of emotional collapse. Therefore, the translator must aim to evoke not only the content but also the emotional and theological purpose of the verse.

Yusuf Ali’s translation, “I only complain of my distraction and anguish to God”, uses the term “distraction” for the Arabic bath-thī, which adds a psychological aspect but risks softening the existential weight of the original phrase. While “anguish” maintains emotional force, the use of “distraction” may lessen the gravity of the sorrow being conveyed. However, the rhythmic and reverent tone of Yusuf Ali’s rendering enhances the sacred quality of the expression, aligning with the verse's functional purpose as a spiritual release through divine address.

Pickthall’s version— “I expose my distress and anguish only unto Allah”—uses stronger diction with “expose,” a verb that connotes emotional vulnerability. This choice increases the intimacy and rawness of the emotional moment. The phrase “only unto Allah” preserves the Qur’anic exclusivity of divine complaint. While “distress and anguish” are accurate emotional terms, the somewhat archaic tone of the phrasing may restrict accessibility for modern readers, though it preserves the gravity of the original.

Sahih International’s translation, “I only complain of my suffering and my grief to Allah”, is clear and contemporary. The terms “suffering” and “grief” are emotionally resonant and accurate, but relatively restrained. While it effectively communicates the intended message and is more accessible to contemporary readers, it offers less rhetorical and devotional force than the earlier translations. Nonetheless, it achieves a balanced level of functional equivalence by preserving the semantic content and spiritual context in straightforward language.

The AI translation, “I only complain of my sorrow and grief to Allah”, is nearly identical to Sahih International’s. While accurate in terms of surface meaning, it lacks emotive depth and poetic modulation. The words “sorrow” and “grief” are clear, but do not convey the intensity of emotional surrender present in the original. Furthermore, the AI model does not enrich or adapt the structure for rhetorical or theological effect; it simply mirrors the input. As such, the functional equivalence is partially achieved, but the emotional resonance and spiritual tone are significantly reduced.

In sum, human translators generally succeed in reflecting the inner tone and devotional function of this deeply emotional verse. While Sahih International and the AI version offer semantic clarity, Yusuf Ali and Pickthall introduce more rhetorical texture and sacred rhythm. The AI translation, though correct, illustrates the limits of non-contextual rendering when dealing with highly affective and spiritually charged expressions of grief in sacred texts.

Technical Commentary on AI Translation

The AI-generated translation of Qur’an, (86) “I only complain of my sorrow and grief to Allah”, demonstrates the model’s preference for semantic clarity and grammatical accuracy. The lexical choices “sorrow” and “grief” are accurate on a denotative level, but they reduce the emotional complexity embedded in the original Arabic. The term “bath-thī” in Qur’anic Arabic refers not just to sorrow but to a sense of emotional dispersion or suppressed anguish that ultimately erupts in divine complaint. GPT-4’s selection of “sorrow” lacks that depth. Additionally, the AI’s rendering lacks any attempt at rhetorical elevation, poetic rhythm, or sacred tone. This is consistent with its training objective: to optimize for fluency and coherence, not spiritual or emotional depth. As in previous cases, the AI’s attention mechanisms focus on matching input-output patterns, without the capacity to engage with the psychological or theological richness of prophetic speech. The resulting translation is emotionally subdued and stylistically plain, thereby diminishing the verse’s power as both lament and affirmation of divine trust.

Synthesis and Reflection

This final example reinforces the pattern observed across the analysis: while AI translations are competent at conveying basic meaning, they lack the interpretive and emotional depth required for sacred texts. In Qur’an (86), the human translators—particularly Pickthall and Yusuf Ali—effectively convey the emotional and spiritual complexity of Ya‘qub’s lament, embedding it in language that evokes both sorrow and divine intimacy. The AI model, by contrast, treats the verse as a neutral statement, failing to replicate its layered effect and rhetorical power. Thus, despite lexical accuracy, the AI translation falls short in delivering functional or hermeneutic equivalence, confirming the ongoing necessity of human insight in the translation of emotionally rich religious discourse.

Conclusion

This study investigated the ways in which the Qur’an conveys key emotional constructs, and how these affective elements are maintained, altered, or attenuated in English translations produced by both human translators and artificial intelligence. To achieve this, the study employed qualitative linguistic approach, analysing a selection of emotionally charged Qur’anic verses and comparing their renderings across multiple English translations. The analysis drew upon key theoretical frameworks, including Steiner’s Hermeneutic Motion and Nida’s Functional Equivalence Theory, to examine how translators negotiate the interplay between semantic fidelity and emotional effect.

The analysis reveals that the Qur’anic concept of “ṣabr” is not merely passive endurance, but an active spiritual posture rooted in faith and trust in God. Other emotional concepts such as “ḥilm,” “raḥma,” and “sakīna” also embody multi-layered theological and psychological meanings that are frequently inadequately captured by direct English equivalents. While human translations exhibit varying degrees of interpretive sensitivity and rhetorical richness, AI-generated translations tend to offer syntactic accuracy at the expense of emotional depth and spiritual resonance.

The study underscores that translating the Qur’an involves more than linguistic transfer, it requires cultural insight, emotional intuition, and interpretive judgment. Qur’anic emotions are not merely descriptive; they are formative, guiding ethical conduct and spiritual growth. As such, translators must go beyond literal rendering and engage in a hermeneutic process that considers the performative and devotional roles of the text.

In the case of AI-generated translations, the limitations are not only emotional but also structural: AI often overlooks intertextual Qur’anic references, fails to account for rhetorical devices such as parallelism and repetition, and cannot situate verses within broader cultural and theological contexts. For example, while a human translator may recognize how the concept of “ṣabr” echoes across multiple surahs to build a spiritual motif, AI treats each verse in isolation, thereby flattening its resonance (Doshi and Sid 2023).(17)

These findings can also be better understood through the lens of translation studies frameworks. Nida’s distinction between dynamic and formal equivalence highlights the tension between literal rendering and functional impact (Nida 1964, 159; Nida and de Waard 1986, 36–39) (18,19), a tension that becomes especially acute in the translation of sacred texts. Similarly, the broader challenges of translating scripture—where linguistic precision must coexist with theological sensitivity—underscore why human mediation remains indispensable in Qur’anic translation (Gutt 2000, 204–206) (20).

In conclusion, this research affirms the necessity of emotionally aware and theologically informed translation practices. Translating emotionally charged Qur’anic expressions requires a careful balance between linguistic precision and spiritual resonance. Future studies might expand the emotional lexicon under investigation, explore reader reception across cultures, or further examine the evolving role of AI in sacred text translation. As AI becomes more involved in translation, ethical boundaries and interdisciplinary methodologies must be considered. Bridging the gap between Arabic source texts and English target texts requires continuous refinement to ensure that the Qur’an’s emotional, spiritual, and ethical dimensions are conveyed with clarity, reverence, and depth.

6. References

1.Pennebaker JW. The secret life of pronouns: What our words say about us. New York: Bloomsbury Press.2011.

 2.Abdel Haleem, M A S.  The Qur’an: A New Translation. Oxford: Oxford University Press. 2005.

3.Izutsu  T.  God and man in the Qur’an: Semantics of the Qur’anic worldview. Kuala Lumpur: Islamic Book Trust.2002

4.Robinson N.  Discovering the Qur’an: A contemporary approach to a veiled text. London: SCM Press. 1996.

5.Rahim IA & Zulkepli M KA A review of Qur’anic rhetoric studies: Techniques and linguistic features. International Journal of Modern Education. 2025; 7(26), 657–674. https://doi.org/ 10. 35631/IJMOE.726044.

6.Ahmad H & Ghafar N A.  Stylistic variation and linguistic strategies in Qur’anic discourse: A rhetorical, phonetic, and translational analysis. Internat J Res. Innovation  Social Sci. 2025;9(4). https://doi.org/ 10. 47772/IJRISS.2025.90400426.

7.Ibn Manzur M Ibn Mukarram. Lisān al-ʿArab. Beirut: Dār Ṣādir.1955.

8.Mir M.  Coherence in the Qur’an: A Study of Islahi’s Concept of Nazm in Tadabbur-i-Qur’an. Indianapolis: Amer Trust Pub.1989.

9.Steiner  G. After Babel: Aspects of Language and Translation. 1st ed. Oxford: Oxford University Press. 1975.

10.Nida E A and Jan D. Waard.  From One Language to Another: Functional Equivalence in Bible Translating. Nashville: Thomas Nelson. 1986.

11.Castilho S, Joss M, Fabio G, Iacer C, John T and Andy W. . “Is Neural Machine Translation the New State of the Art?” Prague Bull Math Lingu 2017;108(1):109–120. https://doi.org/ 10.1515 /pralin-2017-0013.

12.Toral, Antonio, and Andy Way. 2018. “What Level of Quality Can Neural Machine Translation Attain on Literary Text?” In Translation Quality Assessment: From Principles to Practice, vol. 1, edited by Joss Moorkens, Sheila Castilho, Fabio Gaspari, and Sharon Doherty, 263–287. Cham: Springer International Publishing. https://doi.org/10.1007/978-3-319-91241-7_12

14.Abdel Haleem, Muhammad A  S . Understanding the Qur’an: Themes and Style. London: I. B. Tauris.2011ز

15.House, Juliane. 1977. A Model for Translation Quality Assessment. Tübingen: Gunter Narr Verlag.

16.House, Juliane.  Translation Quality Assessment: Past and Present. 2nd ed. London: Routledge.2015. https:// doi. org/10.4324/9781315752837.

17.Doshi A, and Sid C. The Loss in AI Translation: Examining the Pitfalls and Ethical Implications of AI Translation.” Preprint, Research Archive of Rising Scholars.2023. https://doi.org/ 10. 58445/rars.1453.

18.Nida  Eugene A.  Toward a Science of Translating: With Special Reference to Principles and Procedures Involved in Bible Translating. Leiden: Brill.1964.

19.Nida Eugene A and Charles R. Taber.  The Theory and Practice of Translation. Leiden: Brill.1969.

OpenAI. 2024. ChatGPT (GPT-4 Model; Used June–July 2024) [AI Language Model]. San Francisco: OpenAI. https://chat.openai.com.

20.Gutt, Ernst-August. 2000. Translation and Relevance: Cognition and Context. 2nd ed. Manchester: St. Jerome Publishing.