For Nithya. ɛις μνήμην ἡμῶν φιλιας.
What does it mean to be virtuous? One response to this ancient and transcultural question has been genealogical, construing the question as “What has it meant for certain people to be virtuous?” Adopting such a perspectivalism, as I choose to do in this essay, necessarily localizes answers to persons, cultures, communities, and traditions, rooting understandings of virtue in lifeworlds. This essay focuses on a central feature of a world genealogy of virtue—that of adab, an Arabic-Persian word conveying a distinctively Islamic ideal of prestigious sociality. Standing for Hellenic-Islamic norms of urban civility and refinement, adab constituted a historical alternative to Shar‘ia-based models of virtue that were rooted in the verification of Ḥadīs¯ (plural aḥādīs¯), or reports by and on the Prophet Muhammad (Salvatore). A central feature of adab, as this essay bears out, was making sense of texts, a practice I here call literary criticism. Taking premodern literary criticism seriously is, in this sense, a necessary part of taking adab seriously.
This essay argues that premodern Persian literary criticism adopted two overlapping but distinct interpretive dispositions, the charismatic-allegorical and the humanist-philological, and that these two dispositions assumed and furthered two models of friendship, respectively—a vertical kind based on Sufi hierarchy and a horizontal kind based on rhetorical education. It sets forth how these presupposed models of friendship conditioned literary commentary by considering representative passages from three Mughal commentaries on Sa‘dī's گلستان (Gulistān, or The Rose Garden), completed in 1258 in Shiraz. Here, by “friendship” I mean nothing more than a kind of unwritten pact. In the vertical disposition this pact entailed, as I show below, the reader's expectation that the commentator would furnish, by the authority of his charisma and in line with charismatically authorized precedents, allegorical meanings of the text or portion of text at hand. The horizontal disposition entailed the reader's expectation that the commentator would furnish, by the display of empirically falsifiable textual precedent, grammatical-lexicographical-semantic meanings of the text or portion of text at hand. Though involving distinct forms of authority and reading protocols, both kinds of pact entailed trust, and it is only in this basic sense that I call them forms of friendship. Both were forms of affiliation or chosen relation rather than forms of filiation or biological descent or kinship.Footnote 1 This is important because it is affiliation that made these two interpretive methods socially significant rather than private or familial affairs. That is, these affiliative rather than filiative modes of social relation ensured their public and large-scale reality. So basic to these interpretive postures were these forms of friendship or affiliation that literary criticism simply assumed them instead of making them explicit. This is why none of the premodern literary criticism quoted below explicitly cites friendship as its own condition. It does not do so because it did not need to, treating it as axiomatic to its own genre.
And yet, since these notions of sociability are mostly only implicit in these interpretive communities, I make them explicit in two ways. The first way is to note that the Persian commentarial tradition takes for granted the synthesis of Hellenic and Hebraic strands that make up Islamic high culture. That is, on the one hand, it assumes as given the Aristotelian emphasis on the “horizontal” intersubjectivity necessary for the individual's attainment of “virtuous well-being” (Mottahedeh 229–30). For Miskawayh (d. 1030), an inaugural Islamic virtue ethicist, “the perfection of human virtuous well-being can only be realized with friends” (Mottahedeh 231). Urban life was the natural outcome of the individual human need for physical and metaphysical goods accessible only through horizontal webs of mutual services. In The Nasirean Ethics, Naṣīr al-Dīn Ṭūsī (d. 1274) envisions love as a sort of civic affection necessary for elite male attainment of intellectual intimacy with the active intellect, the logically first emanation of God, which makes recourse to legal enforcement mostly unnecessary (Ṭūsī 195–211).
But if theosis or becoming godlike was the overarching aim of Ṭūsī and his many adaptors in the Timurid, Mughal, and Safavid empires, it was because Arabic-Persian imagination of metropolitan camaraderie also assumed a Hebraic inheritance: the necessity or desirability (depending on the genre) of a “vertical” friendship between human and God. As Roy Parviz Mottahedeh writes:
One would expect that in a strongly monotheistic system—and it is difficult to think of a more strongly monotheistic system than Islam—God would be too remote to be a friend. Yet both the Bible and the Qur’ān call Abraham the friend of God. Excepting love affairs, friendship with God or gods would have been unusual in the classical Greek context before Christianity. Despite the strong influence of Greek philosophy, such friendship between God and man is a persistent theme in the Islamic tradition. In later Islamic tradition saintly Sufi mystics are called the awliyā’, the friends of God. (232)
Though this synthesis of Hellenic and Hebraic models of friendship is an undeclared norm in Persian commentaries, we may discern its effects in their institutional settings. This is the second way I explicate these commentaries’ notions of sociability in the course of my analysis.
These two hermeneutic dispositions and their corresponding styles of friendship were developed through an engagement with a small and closed set of Persian literary works that Timurid courtly culture of late-fifteenth-century Herat had canonized. On the evidence of its outsized commentarial reception, Sa‘dī's Gulistān was the most widely read of them. This was not accidental as the very form of the Gulistān, as well as the circumstances of its composition, made it amenable to character formation or ethical education like no other text.
These, then, are the larger political, historical, and literary stakes of taking premodern Persian discourses on literature seriously. But what was premodern literary criticism in Persian? Until the late-nineteenth-century founding of modern universities, literary criticism did not exist under this name or even as a unified field of practices under another name. What did exist was metaliterary discourse in different genres. This is not remarkable. No aesthetic practice lasts without simultaneous feedback. Real-time validations are a necessary condition for any aesthetic practice that is undertaken with the hope of enduring and reaching a wide audience. Persian poetry, a high-prestige linguistic practice that lasted from around the 800s to the 1850s, was no exception. Until the late fifteenth century, its real-time validations took four forms: oral commentary, now mostly lost; anthologization in taz¯kirahs, or variously organized prosopographical anthologies of poetry; dictionaries of poetic lexemes; and poems composed in response to poems. But a written discourse addressed systematically to a literary text from its beginning to end—that is, commentary identifying itself as sharḥ—appeared only in late-fifteenth-century Timurid Herat and soon after in Ottoman courts that modeled Turkish literary norms on Timurid Persian classics (Kuru). As in Ottoman Turkish, sharḥs were also composed in an increasingly locally nested Mughal Indian network between the late sixteenth and mid-nineteenth centuries, though in Persian rather than in a local language.
While مثنوی معنوی (Mas¯navī-i ma‘navī; The Mas¯navī of Inner Significance) by Rūmī (d. 1273), پند نامه (Pandnāmah Footnote 2; Book of Advice) by Farīd al-Dīn ‘Aṭṭār (d. 1221), and the دیوان (Dīvān; Collected Poems) of Ḥāfiẓ (d. ca. 1390) were all topics of commentaries from the late fifteenth century onward across the Persianate world, it was the Gulistān that was bound up with a child's education and thus with ethically oriented commentary addressed to readers of every age. Explaining why this was the case also allows us to understand why the commentarial tradition on this text was characterized by a tension between the two aforementioned models of friendship. In the following section I argue that, while Sa‘dī's oeuvre as a whole was characterized by both forms of friendship, it was the uneven synthesis of the two in the Gulistān that made it so amenable to education or ethical formation and thus also the proving ground for commentaries based on one or the other forms of friendship.
Inscribed into the speech situation in which the Gulistān locates its own origin is a model of egalitarian male friendship. The work belongs to the premodern Persian genre of akhlāq, or ethics, a genre addressing advice on topics of state, home, and character formation to a Muslim king or his functional equivalent. All the Arabic and Persian texts in this genre are related through ever-greater degrees of elaboration to Aristotle's Nicomachean Ethics (Mottahedeh 235). Despite its dissimilarities in form and format with the more systematic texts in this genre, it shares with them its pervasive emphasis on cautious moderation in the interhuman exchange of words, goods, and gifts. Accordingly, the God of the Gulistān, unlike the God of Sufi allegories like ‘Aṭṭār's منطق الطیر (Manṭiq al-ṭayr; The Conference of the Birds), is an overseer of equitable human exchanges. We may get at this model by considering Sa‘dī's account in his much-imitated and oft-memorized preface to the Gulistān of how he came to compose the text.
Despairing of his “عمر تلف کرده ” (“wasted life”; Sa‘dī, Kulliyāt-i Sa‘dī 25; Sa‘dī, Gulistan 5), the narrator resolves to fall silent. An old and intimate friend visits him “به رسم قدیم ” (“as was his wont”; 26; 5). Finding he will not speak, the friend learns of Sa‘dī's resolution from his attendant and appeals to him thus: “آزردن دوستان جهلست و کفارت یمین سهل ” (“To hurt one's friends is an act of ignorance and to atone for breaking an oath is easy”; 26–27; 6). The friend refuses to leave until Sa‘dī addresses him “بر عادت مألوف و طریق معروف ” (“as was his custom”; 26; 6). He says to Sa‘dī:
Throughout the Gulistān words, like goods and gifts, are worth nothing if not exchanged. It is how, when, and where they are exchanged that the text attempts to judge. Keeping one's silence when one should speak exposes one to the risk of fundamental misrecognition. True character shines out in verbal mastery, not mere appearances, giving felicitous social evidence that one is dominated by the nafs-i nāṭiqa (“rational part of the soul”) that Aristotle, foundational to Islamic ethics, considered the trait that differentiated human from nonhuman animals. He finds himself unable to keep his silence, for he
زبان از مکالمه ٔ او در کشیدن قوّت نداشتم و روی از محاوره ٔ او گردانیدن مروّت ندانستم که یار موافق بود و ارادت صادق . (Sa‘dī, Kulliyāt-i Sa‘dī 27)
did not consider it virtuous to refrain from conversing with him, for he was an agreeable friend and sincerely amicable. (Sa‘dī, Gulistan 6)
Sa‘dī addresses him and they walk out together into another friend's garden as it is blooming in spring. After a night in the garden's orchard, the friend gathers flowers in his skirts as they prepare to return. But Sa‘dī reminds him:
گل بستان را چنان که دانی بقایی و عهد گلستان را وفایی نباشد؛ و حکما گفته اند هر چه نپاید دلبستگی را نشاید. گفتا طریق چیست؟ گفتم برای نزهت ناظران و فسحت حاضران کتاب گلستان توانم تصنیف کردن که باد خزان را بر ورق او دست تطاول نباشد و گردش زمان عیش ربیعش را به طیش خریف مبدل نکند . (28–27)
“Orchard flowers, as you know, do not last and the garden has no fidelity. The wise have said, ‘What does not last is not worthy of attachment.’”
“What is to be done?” he asked.
“I can compose, for the delight of spectators, and the enjoyment of those present, a book called Rose Garden, upon the leaves of which the chill wind of autumn will make no inroads and the springtime harmony of which the vicissitudes of time will never transform into the stridency of autumn.” (7)
Friendship, then, is not only the exchange of language in honor of ethical commitments that antedate any individual decision; it is also the very condition under which Sa‘dī can assume authorship. Moreover, it is authorship of a verbal and therefore potentially immortal imitation of a creaturely and thus mortal garden. Literary immortality had always been a motivation for authorship in Persian literary culture. Sa‘dī was paradigmatic in making egalitarian friendship a condition for it.
But in fact Sa‘dī's preface condenses into itself both models of friendship. Sa‘dī composed this work after a period away from Shiraz and arguably to reestablish his relations with the court. Accordingly, he dedicated it to the king, Sa‘d Abū Bakr Sa‘d bin Zangī, in honor of whom Sa‘dī chose his own pen name. The dedicatory passage appears immediately after the foregoing passage on his own friend and prefaces the king's proper name with a sequence of Persian and Arabic titles that cite in prose what would have been a familiar panegyric topos:
و تمام آنگه شود به حقیقت که پسندیده آید در بارگاه شاه جهان پناه سایه ٔ کردگار و پرتو لطف پروردگار ذخر زمان و کهف امان (28)
But in truth it will be finished when it is approved and read with lordly kindness at the court of the king of the earth, asylum of the world, shadow of the Creator and ray of the Provider's kindness, treasure house of the age and cave of safety. (7)
In a panegyric, such a sequence of titles would signal the transition from the tableau of the exordium (naṣīb) to praise (madḥ) of the dedicatee. By invoking that topos here, Sa‘dī signals that the foregoing passage, which we read as a reflection on egalitarian friendship, was a prose simulation of an exordium tableau or mise-en-scène with which he now breaks to rise to a more privileged relationship: a hierarchical one. Though this is not a Sufi hierarchy of the sort I discuss below, it shares with it a structure of servitude and patronage. Just as his ghazals ambiguously mirrored the poet's own relations with his patron in their lover-speaker's relations with his beloved (Meisami 277–79), so too does Sa‘dī as narrator of the Gulistān eroticize his text:
دیگر عروس فکر من از بی جمالی سر بر نیارد و دیده یأس از پشت پای خجالت بر ندارد و در زمره صاحبدلان متجلی نشود مگر آن گه که متحلّی گردد به زیور قبول امیر کبیر (28)
Furthermore, the bride of my thought is too unlovely to lift her head and too despairing to raise her eyes and come out from behind her bashfulness and appear in the ranks of persons of understanding unless she is bedecked by the finery of acceptance by the great and victorious prince. (8)
As in the poetry of Niẓāmī (d. 1209) and Khāqānī (d. 1190) before him, model reading was epithalamic, a consummation by an active groom reader of a stereotypically passive bridal text.Footnote 3
As in the preface to the Gulistān, so in Sa‘dī's ghazal corpus, the friend in fact ranges between his symmetrical and asymmetrical identities, occasionally appearing as a member of a horizontal camaraderie, as it were, and mostly appearing vertically in an erotic hierarchy. Some ghazals speak of the friend in an egalitarian sense as a stay or anchor in a world fraught with accident and as a guarantee in a world of risky transactions while most eroticize or lament his cruel distance and unattainability. It is this ambiguity that Mana Kia's otherwise insightful essay papers over when it contrasts a monolithic Persianate sense of selfhood based on loving mutuality (presented, problematically, as stable since Sa‘dī) with an equally monolithic one based on alterity and conflict that took shape in the European Enlightenment:
In these articulations of an autonomous Enlightenment self as collective category, self-definition is predicated on a mutually exclusive sense of difference. This form of self-definition exists in Persian modernist writings as well. But simultaneously, and perhaps more important, up through the late nineteenth century (at least), it is the beloved friend through whom a modern and increasingly Iranian self was defined, a self based on intimacy, commonality, and similarity. (“Indian Friends” 399)
Not only does this account ignore Edmund Husserl's early-twentieth-century “European” discovery of the other within a transcendental phenomenology of selfhood (124, 148), it also ignores the salience to Islamic formulations of selfhood of the kind of xenology represented by the medieval genre of ‘ajāyib or mirabilia.Footnote 4
Returning briefly, then, to the vertical model of friendship in Sa‘dī, the abject pleading of the friend or lover (yār and dūst convey both “friend” and “beloved” in Persian) and his beloved's heedless cruelty form the ghazal's arch trope, and Sa‘dī's ghazals exult in this masochistic aesthetic. Here, too, friendship constitutes selfhood. But it does so through an eroticized difference and hierarchy and characterizes the human relationship with God in Sa‘dī's بوستان (Būstān; The Orchard). It also appears in some of the anecdotes in the fifth chapter of the Gulistān, which comprises anecdotes on love and youth. Friendship as sacralized erotic hierarchy forms the central topic of Aḥmad al-Ghazālī's prosimetrum سوانح (Sawāniḥ; Intimations [c. 1114]), the earliest known Persian-language exposition of the love mysticism of ninth-through-twelfth-century Khurasan (Pourjavady, “Ḡazālī”). Through a close comparison of the phrases and syntax of Sawāniḥ and Sa‘dī's ghazals, Nasrollah Pourjavady has shown that most of the latter are explorations and elaborations of Aḥmad al-Ghazālī's theology of love (“Sa‘dī”). Though it began as a secular and courtly mode of love, this erotic hierarchy had thus come to be appropriated by Sufis well before Sa‘dī's time, an appropriation that entailed the allegorization of the lover and beloved in terms of oppositions and as human and divine respectively. But it is not central to the Gulistān because it would impede the kind of circulation (of words, goods, and gifts) Sa‘dī is calling for. This is why, when Sa‘dī recognizably drew some of his anecdotes from Sawāniḥ, he stripped them of their mystical-allegorical dimension. Persisting in a human hierarchy where the weaker party gives but gets nothing in return would be antithetical to the spirit of cautious moderation and constant circulation of goods, words, and gifts that characterizes akhlāq discourse. Sa‘dī consequently restricts its appearances even as he suggests by the identity of the plots of some anecdotes across his akhlāq and mystical works that the former is only a modulation—perhaps even diminishment—of the latter.
And yet, it is a tension between precisely these two models of friendship that characterizes the commentarial tradition on the Gulistān. Between 1494, when ‘Uways bin ‘Alā al-Dīn Ādam glossed the text for Bahmani Sultan, and Ghiyās¯ al-Dīn Rāmpūrī's 1843 commentary on it for a prince of Rampur, India alone accounts for at least a dozen Persian-language commentaries on the Gulistān (Lewis). This essay focuses on three from the late phase of this commentarial tradition because they assimilate earlier Mughal commentaries by engaging them; and because they serve as evidence for the distinction between charismatic-allegorical and humanist-philological hermeneutic postures and the corresponding forms of friendship these assume. That this commentarial tradition is bookended by commentaries addressed to princes is a sign that early modern Islamic courtly culture was a thorough synthesis of the aforementioned Islamic-Hellenic and Islamic-Hebraic models of friendship.
A prefatory note: it will appear that whereas the foregoing interpretation of Sa‘dī's text was concerned with how the social practice of friendship interacts with the practice of authorship, the following interpretations of the commentaries concern hermeneutical practices more than social ones. But, as we have observed above, even this social practice was in fact a stylized literary image of egalitarian friendship mediated by the genre convention of the nasīb, or opening tableau of a panegyric. In this sense, all the discussions of reading practices here are, to some extent, discussions of idealized images.
The following pages mainly focus on three commentaries: Sirāj al-Dīn Khān Ārzū's خیابان : شرح گلستان سعدی (Khiyābān: Sharḥ-i Gulistān-i Sa‘dī; The Garden Pathway: A Commentary on Sa‘dī's Rose Garden [Delhi, 1750s]), Valī Muhammad Akbarābādī's شرح گلستان سعدی (Sharḥ-i Gulistān-i Sa‘dī; A Commentary on Sa‘dī's Rose Garden [Agra, 1753]), and Ghiyās¯ al-Dīn Rāmpūrī's بهار باران : شرح گلستان سعدی (Bahār-i Bārān: Sharḥ-i Gulistān-i Sa‘dī; A Springtime of Rain: A Commentary on Sa‘dī's Rose Garden [Rampur, 1843]). These form part of what Eric Van Lit would call “the restricted commentary tradition” of the Gulistān in that their structure corresponds from start to finish to that of Sa‘dī's source text; and in that they “form the backbone of the commentary tradition. They are the most widely circulated and read texts of this set, and thus the most influential.” This therefore does not include the countless freestanding marginal notes to the source text or partial or complete prosimetric imitations of the Gulistān or the uncountable grafts of parts of its text onto texts by others—all forming the wider “commentary tradition” set within the still wider “textual tradition” made up of Sa‘dī's source text and the sum total of hypertexts consciously referring to it. However, we may assume that all of the latter kinds of text were variously aware of the restricted commentary tradition we consider.
Sirāj al-Dīn ‘Alī Khān Ārzū
An overarching thesis traverses Ārzū's oeuvre: “that the trans-temporal and trans-spatial concept and criterion of “linguistic purity” (faṣāḥat) had always had local content specific to pedagogically trained peoples within a period and across periods” (Keshavmurthy, “Ḵhān-e Ārezu”).Footnote 5 Each of his works bears out this thesis in genre-specific ways: his Dīvāns or collections of Persian poetry innovate on the compound metaphors of poets he names or cites himself; his commentaries on works from the Persian poetic canon assume that the meanings of words and phrases drift over time and thus need commentary; his treatises staking out his positions on debates over apposite poetic style all make cases for the period-specific and training-based nature of eloquence; his dictionaries of poetic idioms distinguish period-wise between such idioms; and his four other treatises on metaphor making and syntax in the Persian ghazal and the history of the Persian language equally assume that textual meaning is subject to shifts over time that could be controlled, if only partially, by prescriptive philology like his own. No part of this multigeneric project has any place for the primordial substitutability of allegorical meaning among Sufis where the ghazal beloved's locks of hair have always stood for God's infinite attributes. Nothing of his historicist philology would have been possible without the rigorously falsifiable evidentiary protocol he called taḥqīq, a protocol that assumed humanist erudition in the Persian poetic canon of the sort that anyone could acquire in theory. In this sense, his philology was egalitarian in its methodological assumptions.
This egalitarianism served and expressed its institutional setting. Ārzū inhabited a niche in the Mughal bureaucracy, the largest ever in India. Such imperial expansion required increasing numbers of bureaucrats with something less than the sort of polymathic expertise traditionally acquired by apprenticeship to a teacher. Early modern commentary and philology like Ārzū's was one of the means by which an aspiring bureaucrat might train himself in the Persian literary canon where he would previously have had to seek out a living teacher. Nile Green characterizes this as the late Mughal displacement of epistemic authority from “anthropocentric” sources to “bibliocentric” ones (Green). Such bibliocentrism, as evidenced in the dense intertextuality of the urban Persian literary culture of Mughal India and the centrality to its bureaucracy of Ṭūsī's Nasirean Ethics and its adaptations, drew on the Islamic-Hellenic understanding of urban sociality this essay began by elaborating.Footnote 6
Here is a passage from Ārzū's commentary. The passage glosses only one word in Sa‘dī's text—“chess”:
قوله : شطرنج بازی معروف؛ در اصل اگر لفظ هندی است چنانکه از کتب تواریخ معلوم می شود پس معرب
چترنک بود وآن بازی است که در مشرق اوهند که بدین زبان پورب گویند شهرت دارد و اگر فارسی بود پس معرب سترنک خواهد بود. چنانکه در جهانگیری است و سترنک گیاهی است که به صورت آدمی باشد و در چین روید و چون در بازی مذکور نیز نام انسان بر چوب تراشیده مثل شاه و وزیر بگزارند بدین نام موسوم شده . بر هر تقدیر در اصل به فتح است و معرب آن به کسر اول چنانکه در قاموس گفته . (155)
He [Sa‘dī] says, “Chess” [shaṭranj]: A well-known game; if originally an Indian [hindī] word, as appears to be the case from historical texts, then the form it assumes when adopted into Arabic [mu‘arrab] would be chaturang, which is a game that in the east of Avadh—called Purab in this language—is popular. If Persian [in origin], the form it assumes when adopted would be satrang, as noted in the Jahāngīrī. Satrang is a man-shaped plant [the mandrake] that grows in China and since in the game in question, too, they carve human names like “king” and “minister” onto pieces of wood, it was called by this name. At any rate, it is originally pronounced with the vowel “a” [fatḥ] and the form it assumes when adopted into Arabic vocalizes the first syllable with the short vowel ‘i’, as noted in the Qāmūs.
As with many such glosses in Ārzū's commentary, we wonder why he would have felt the need to gloss a term whose referent he himself says is “well-known.” Apparently at issue was the word's shape across languages rather than its sense. But surely this was equally true of any other noun in Sa‘dī's text, so why did he not gloss every such noun? The answer lies in the famously international origins of chess. Its ambivalently Indian and Persian origins determine an ambivalence in its nominal form in Arabic. And since the form it took based on its putative Persian origin happened to coincide with the name of the mandrake plant to which its pieces bore a physical resemblance, its Persian name became satrang.Footnote 7 Chess was the perfect pretext for Ārzū to offer a miniature lesson in his historicist philology. For even this familiar noun, he was implying, bore traces of the transimperial multilingualism that had constituted Persian. His مثمر (Mus¯mir; Fruition), the most conceptually articulated and extensive of his treatises on Persian, dedicates special attention to the rule-governed transformation of words as they travel from one language to another, especially across Arabic, Persian, and Hindavi. Here, being of Indian or Persian origin accounts for two distinct but equally valid vocalizations: shatranj in Persian, as per its original pronunciation, and shitranj in Arabic, as per rules of lexical adoption laid down in al-Firuzābādī's famous Arabic dictionary of 1410, القاموس المحيط (al-Qāmūs al-muḥīṭ; The Encircling Ocean).
Even as brief an entry as this one contains traces of Ārzū's core theoretical-philological preoccupation: the simultaneously rule-governed and aleatory character of literary language and the authority to decide which rules would determine what historical accident had thrown at a language and how. The citations of authorities such as القاموس (al-Qāmūs; The Ocean) and the فرهنگ جهانگیری (Farhang-i Jahāngīrī; Jahāngīr's Dictionary), as well as personal knowledge of geographically specific Hindavi vocabulary, are performances of the kind of philological competence that Ārzū believed authorized eloquence and negated the criteria of ethnicity, climate, and place of origin that his Iranian rival in Delhi, Shaykh ‘Alī Ḥazīn, had invoked in his resentful loss of home in the courtly circles of Safavid Isfahan (Kia, “Accounting”). What is key here is that none of these criteria assume charismatic or nonrational forms of authority transmitted by revelation and Sufi initiation. Rather, they assume criteria that are falsifiable by counterevidence of the same linguistic kind. In this sense, Ārzū's expectation that his readers would be persuaded on the grounds of what he presents in his preface as his proven skills in literary commentary and thus on grounds of secular trust presupposes the horizontal form of friendship or affiliation this essay opened by explaining.
But the Gulistān was the ideal text for such a demonstration because it was the paradigm of eloquence; and because its key concern was to model sociality at a time of political upheaval and cultural flux. As noted, Sa‘dī composed the text to reestablish his connections with the Mongol court at Shiraz after a period away from the city. He did so in 1258, the very year in which the Mongols ransacked Abbasid Baghdad and killed the last Abbasid caliph. Not surprisingly, his overarching conception of human sociality and arguably of history itself was that it was fraught with risk and could not be assimilated into rules that made it safely predictable. Concerning himself with the skills necessary for a safe, well-provisioned, and pious life, he advocated an overall attitude of cautious moderation in all matters. His eight chapters convey something closer to rough-and-ready working principles through rhetorically pleasing and thus imitable examples than to rules in the abstract.
This is why the bulk of his text, though loosely ordered into eight thematic chapters, is made up of anecdotes. Typically short, discrete narratives, these anecdotes illustrate the larger theme and conclude with lines of verse that abstract and eternalize what the narrative presents in its concrete and passing particulars. This was, as Pourjavady has suggested (“Sa‘dī”), a format he likely borrowed from Aḥmad al-Ghazālī's aforementioned Persian prosimetrum, Sawāniḥ. As such, the Gulistān may be contrasted with the systematicity of such classic works in the akhlāq genre as Naṣīr al-Dīn Ṭūsī's Nasirean Ethics and its many adaptations that derive their overall sequence and contents from Arabic adaptations of Aristotle's Nicomachean Ethics.Footnote 8 None of these works came to be abidingly associated with character formation as the Gulistān did and none of them, relatedly, were continuously commented on since they were composed.Footnote 9
Without making explicit the horizontal style of friendship its humanist-philological hermeneutics presupposed, Ārzū's commentary on the Gulistān exploited the text's muted interest in the Sufi metaphysical, theological, and philosophical frames through which Sa‘dī could have chosen to teach the good life. Instead, it was the secular and egalitarian form of friendship the text and Mughal bureaucratic bibliocentrism assumed as well as its fragmented anecdotal character that Ārzū exploited as a pretext to pursue his own interests in philology and its politics.
Valī Muhammad Akbarābādī
The salient feature of Valī Muhammad Akbarābādī's voluminous 1753 commentary on the Gulistān is its Sufi frame of reference. Without any overt justification for it, Akbarābādī opens his text in an allegorical mode that he sustains till the end, a mode that expects that the reader will give credence to a Sufi metaphysics already prestigious by Mughal political appropriations of it. This expectation was also bolstered by Akbarābādī's previously successful use of this metaphysics to interpret Rūmī (d. 1273). He composed this commentary after he completed a seven-volume commentary on Rūmī's Mas¯navī-i ma‘navī, arguably the most famous text in the canon of Persian-language Sufism. His approach to Sa‘dī's text is continuous with the one he takes toward Rumī, whom he reads as sharing a metaphysics with Ibn ‘Arabī (d. 1240).Footnote 10 Specifically, he invokes the theistic monism of Ibn ‘Arabī, the most influential thinker in the history of Sufism, through the copious use of his technical terminology.Footnote 11 He combines this with a metaphysics of light drawn from the illuminationist tradition of Suhrawardī (d. 1191), also a widely influential thinker in the tradition of philosophical Sufism.Footnote 12 What strikes any modern reader familiar with Sa‘dī's plainly non-Sufi text is how Akbarābādī manages to read the text in such sacralizing ways.
His key method to this end is lemmatization. He isolates lines, sentences, and phrases from their surrounding textual context. This allows him to ignore the formulaic quality and arguably secondary importance of Sa‘dī's uses of theological-philosophical terms and to amplify what he presents as their concealed and specialized senses. It is as if he were reading with eyes peeled for signs he could read as coded references to Ibn ‘Arabī and Suhrawardī's ideas. We will confine ourselves to the following example taken from Akbarābādī's commentary on these lines from Sa‘dī's famous preface, quoted in part in the beginning by Akbarābādī but which I complete in square brackets, using Wheeler Thackston's translation:
قوله ابر و باد و مه و خورشید و فلک در کاراند + الی البیتین مدلول لفظی این هر دو بیت درغایت وضوح است لیکن مقصود در غایت خفاست چه فهمیدن آن موقوف بر علم توحید و تحقیق است و فقیر بجهت طالبان تحقیق به اجمال بیان میکند اول سخنی چند که موقوف علیه بیان مقصود است باید گفت تا تفهیم مقصود آسان شود پس بشنو که ذات حق که هستی مطلق ست جز در تعینات و تقیدات که مظاهر اسما اند بلکه عین اسما در خارج موجود نیست...و چون این مقدمات در یافتی پس بدان که در بیت اول و مصرع ثانی بیت دوم اشارت بمرتبه ربوبیت و معبودیت ابر و باد و ماه و خورشید ست و بمرتبه مربوبیت و عبودیت انسان و در مصرع اول بیت دوم اشارت بمرتبه مربوبیت و عبدیت امور مذکوره است و بمرتبه ربوبیت و معبودیت انسان
(Sharḥ-i Gulistān 9)
He [Sa‘dī] says, “Cloud, wind, moon, sun and celestial sphere—all are at work [so that / You may obtain a loaf of bread and not eat it in heedlessness. / All their heads are spinning and they are obedient: / It is not fair for you not to be obedient” (Sa‘dī, Kulliyāt-i Sa‘dī 23; Sa‘dī, Gulistan 1)]—The literal referent of both these couplets is extremely clear but their intended sense is extremely hidden for understanding it depends on knowledge of monotheism and verification. This humble author will explain it in brief for students of verification. First, a few words are in order to explain the purpose so as to ease its comprehension. Know, therefore, that the substance of the Truth [z¯āt-i ḥaqq] that is absolute existence is not present externally except as actualizations and delimitations that are manifestations of the Names, indeed are the Names themselves. … When you have understood these preliminaries, know that the first distich and the second hemistich of the second distich are an allusion to the rank of the lordship and being-worshipped of clouds, wind, moon, and sun and the rank of worshipper and servant-hood of the human; and the first hemistich of the second distich is an allusion to the worshipper-hood and servant-hood of the aforementioned actions and to the rank of the lordship and being-worshipped of the human.
Anyone familiar with Ibn ‘Arabī's monism would recognize Akbarābādī's description of God as “absolute existence,” the cosmogony that takes the form of the actualization of God's names, and the essential ambivalence that inscribes divinity and creature-hood into each such actualization. But anyone familiar with Sa‘dī's preface would also recognize Sa‘dī's two distichs as innocent in their plain sense of any allusion to Ibn ‘Arabī and as a familiar topos in such prefaces that opened by praising God in prose and verse before moving on to praise of His last prophet. This is why Akbarābādī speaks of their “literal referent” being “extremely clear” and their “intended sense” being “extremely hidden.”
But how does Akbarābādī know what Sa‘dī actually meant? He does not tell us; he simply performs his Sufi authority by his use of ideas and terms from the tradition of Ibn ‘Arabī. In this sense, his commentary assumes that the reader is an initiate to this style of Sufi hermeneutics. The modern scholar Najīb Māyil Haravī says that one of the reasons for the popularity of Akbarābādī's commentaries was his flowing prose, which “bears little resemblance to the madrasa variety of prose of the subcontinent but is connected to the mystical variety of prose of that region” (Haravī, ye-be). Although we still know almost nothing of Akbarābādī's life, this suggests he inhabited a social space that synthesized Persian-language pedagogy and mysticism. Such a synthesis already characterized the moderate, office-bearer's Sufi asceticism of the Mughal bureaucracy by the mid–seventeenth century, allowing commentators the choice of varying combinations of charismatic allegory and humanist philology.Footnote 13 Akbarābādī was among those who chose to foreground the former.
This choice assumed an initiated reader, an assumption he shares with an inaugural exercise in Persian literary allegory in Ibn ‘Arabī's tradition, Maḥmūd Shabistarī's famous گلشن راز (Gulshan-i rāz; The Rose Garden of Secrecy [1317, Tabriz]), a short verse encyclopedia summarizing answers to seventeen questions on Sufi topics posed to him by a visiting scholar. One of these answers was in response to the very question I ask above: how do Sufi allegorists or “the people of inner meaning” who inhabit communally recognized stations and states of intimacy with or distance from God understand formulations of worldly attachment and desire like the beloved's “countenance and locks” and his “cheek-down and mole” (Shabistarī 97)? Shabistarī insists that his Sufi readers only use such terms of physical description if they have experienced one of three states (ḥālat): “an erasure of the self's contingent attributes when it is overcome by the experience of God's essence” (fanā), “a bewildered suspension of reason in witnessing the divine beloved ” (sukr), or “a helpless witnessing of divinely disclosed sensory images corresponding to super-sensory realities” (dalāl [97]). Note that although Shabistarī applied rules to Sufi poetry that celebrated God-intoxicated rule-breaking, it was a communally recognized inner emotional state prior to all rules that justified the rule-governed practice of allegory. It was affect—in the sense of a transindividual emotion—that later validated allegory such as Akbarābādī's.
Who had the authority to identify such an affect and decide on its validity? It was the Sufi preceptor, but not just any Sufi preceptor. For the proper name Ibn ‘Arabī, or “Akbarian” from his title Shaykh al-akbar, or “the greatest Shaykh,” stood for what became a comprehensive way of speaking of reality. Practitioners of this discourse practiced a return to his texts and to texts that returned to them, renewing the tradition on his inaugural terms. Invoking Ibn ‘Arabī as such an “initiator of a discursive practice,” Akbarābādī would have identified himself as a Sufi allegorist trained in the discursive rules of and authorized by Ibn ‘Arabī, who called himself khatm al-awliyā, or “the Seal of the Saints” (Foucault 132). Walāya, sometimes translated as “sainthood,” also means “friendship.”
For a sense of this notion of friendship in Ibn ‘Arabī, here is William Chittick:
“Friendship” [walāya] derives from the Quranic term “friend” (wali). This Arabic word has a range of meanings, any or all of which may be meant when it is used: friend, someone close, someone given authority, benefactor, protector. The Quran makes it one of God's names, and it also speaks of God's friends and the friends of Satan. The friends of God are those who he has brought near himself, those who he protects, and those to whom, on the basis of their special closeness, he has given a certain authority and rulership, if only over their own egocentric tendencies. (Ibn ‘Arabi 11)
Most of the senses Chittick explicates imply a hierarchy: “someone given authority, benefactor, protector.” That walī is also one of God's names is itself a sign that we are not speaking here of an egalitarian friendship. Notice, too, that God distinguishes his friends by granting them a certain authority. It is on this authority that Akbarābādī could coordinate an already formed cosmogonic narrative and its set of cosmogonic terms with a passage of Sa‘dī's verse whose overt sense suggests nothing of that narrative or its terminology. We can expect that Akbarābādī's Sufi readers who shared his Akbarian-Ishrāqī beliefs, would have recognized this narrative and validated his coordination of it with Sa‘dī.Footnote 14
Akbarābādī does this repeatedly whenever he can. When he cannot, he speaks in the voice of the encyclopedic lexicographer, the meticulous philologist, and the speculative theologian (mutakallim). These shifting voices testify to the cross-pollination of the rational and transmitted disciplines in late Mughal madrasa syllabi. But they are held together by his most recurrent voice, which is that of an allegorist in Ibn ‘Arabī's tradition and thus one that invokes walāya, the sacralized and hierarchical model of friendship rather than the egalitarian one this essay opened by discussing.
Ghiyās¯ al-Dīn Rāmpūrī
Even knowing him only for his famous 1827 dictionary, غياث اللغات (Ghiyās¯ al-lughāt; A Guide to Vocabulary), we can infer that Ghiyās¯ al-Dīn Rāmpūrī (ca. 1785–1852) assumed horizontal or egalitarian friendship in his commentary on the Gulistān, for lexicography, a discipline ancillary to poetry in Persian, was committed to mapping the nonsystematic dispersal of poetic phrases and idioms and was thus inimical to allegory. Perhaps even his career, mostly spent as a tutor in Persian to the kings and princes of the little North Indian kingdom of Rampur, explains his schoolmasterly commitment to the humanist-philological mode of literary interpretation.Footnote 15 The philological and egalitarian presuppositions of his commentary were also arguably important to the verifiability and longevity of his authority in the Urdu literary culture of his milieu,Footnote 16 for Sufi allegory requires that the reader take its metaphysics for granted in order to give credence to its glosses. But it is not clear that such metaphysics was pedagogically useful for the practice of poetic oversight (islāḥ), whereby teachers such as Rāmpūrī edited their students’ ghazals. Professionally embedded in his milieu as a Persian-language teacher rather than a Sufi, Rāmpūrī followed Ārzū, despite the official British replacement in 1837 of Persian by any other language and script in the Bengal Code, in adopting a philological persona and the egalitarian disciplinary protocols this assumed.Footnote 17
These contexts suffice to explain Rāmpūrī's dominant persona in this commentary (as in his other commentaries and prefaces)—that of a Persian-language teacher and instructor to novices in the Islamic rational and transmitted disciplines who had no particular philosophical or Sufi agenda of his own; that of an encyclopedic generalist.Footnote 18 Here are the beginning and the end of his gloss on the opening lines of Sa‘dī's preface. Note that here, as frequently elsewhere in his commentary, he includes an addendum under “Benefit”:
هر نفسی که فرو میرود ممد حیات است و چون بر می آید مفرح ذات پس در هر نفسی دو نعمت موجود و بر هر نعمتی شکری واجب ش نفس بفتحین دم ممد بضم میم و کسر میم ثانی و تشدید دال که بکسر اضافت مکسور است بمعنی درازی دهنده و مدد کننده … فایده مخفی نماند که هر دم بدو حرکت یعنی فرو رفتن و بالا بر آمدن تمام میشود برای آنکه چون مزاج قلب و روح نهایت گرم است اگر اصلاح حرارتش نشود حیات بر حیوان عند العقل بس دشوار … واجب و لازم و ضرور بمعنی عبادتی که ادایش چون فرض لازم باشد مگر منکر اعتقادش کافر نگردد بخلاف فرض که منکر آن کافر گردد (5)
He [Sa‘dī] says, “Every breath that goes in is an extension of life; and when it comes out it is a relief to the individual. Therefore in every breath there are two benefits, and for each and every benefit thanks are necessary [vājib].” Commentator: Nafas with two a vowel signs—“breath.” Mumidd with the u vowel sign on mīm, the i vowel sign on the second mīm and a geminated dāl that, with the i vowel sign, is a consonant marked with a vowel meaning “prolonger” and “helper.” … Benefit: Let it not be hidden that each breath is completed in two movements—descending and ascending. Since the temperament of the heart and soul is extremely hot, if its heat is not relieved life becomes very difficult for a rational animal. … Necessary [vājib], obligatory [lāzim], and required [zarūr]—meaning acts of worship whose performance, like that of “religious duty” [farẓ], is obligatory but nonbelief in which does not make the denier an infidel in contrast to “religious duty,” the denial of which makes a person an infidel.
Rāmpūrī, the punctilious lexicographer who also studied, practiced, and wrote on Islamic medicine, offers a simple explanation in Galenic terms of hot and cold air in the arteries and veins for why Sa‘dī would say every inhalation “extends life” and every exhalation “is a relief to the individual.” Notably, he includes a detail local to his Indian milieu, where Sufis and Islamic physicians had long incorporated yogic breath training into their spiritual-medical practices. In the same passage (in lines omitted here for reasons of space) he distinguishes, like Akbarābādī, between two kinds of ye (the indefinite article as well as the singular marker) and also implicitly takes issue with Akbarābādī, who distinguishes like him between two apparent synonyms—vājib and farẓ—but fails to notice that they are not in fact synonyms because the former refers to a duty nonbelief in which does not make you an unbeliever unlike the nonbelief in the latter.
The overall interpretive spirit here is to demonstrate the scientific, grammatical, rhetorical, and juridical appositeness of Sa‘dī's lexical choices. But one does not get the sense, as one does with Akbarābādī, that Rāmpūrī is using Sa‘dī's text as pretext for his own metaphysical agenda. Rather, he mainly wants to explain Sa‘dī's text and confirm its judiciousness, not appropriate it to advocate his own agenda. This is why his glosses do not seem incongruent with what goes before and after the glossed passage. Often, of such multidisciplinary glosses it is possible to say Rāmpūrī wants to give his students a joint lesson in theology, Arabic grammar, and pronunciation by way of Sa‘dī's Qur'an quotation (Rāmpūrī 7)Footnote 19 and a simultaneous lesson in juridical differences between Shi‘i and Sunni attitudes toward the Prophet Muhammad's family and in the differences between Arabic and Persian prosody by glossing a ḥadīs¯ (13). But it seems equally probable that Sa‘dī either did indeed intend such a theological, juridical, or grammatical point or at least would not have disagreed with Rāmpūrī's gloss of it. No special access to an esoteric core of meaning distinguishes Rāmpūrī's glosses. He pledged formal allegiance to a Sufi saint in Rampur but did not draw on this relationship to authorize esoteric interpretations of Sa‘dī's text. In this sense, Rāmpūrī's hermeneutic sobriety returns to Ārzū's dependence on the plain sense of Sa‘dī's text and the egalitarian or horizontal model of friendship it invokes.
This essay opened by briefly setting the foregoing analyses in the context of Islamic virtue ethics. It seems appropriate to conclude it by circling back to the question of virtue to ask what the stakes might be for the humanities of taking these Persian commentaries seriously. The widest cross-cultural frame of reference for adab remains that of the American historian Marshal Hodgson. When, in the 1960s, he wrote his still-authoritative three-volume account of Islam as a phenomenon of world history, The Venture of Islam (published in 1975–77), his principle of unity for his project was a moral one. He understood Islam to be a phenomenon of world-historical significance not just for Muslims but for all human societies that were legatees of the Axial Age, when, between 800 and 200 BCE, the world's first religions of salvation emerged. These religions—Judaism, Mazdeism, Buddhism, and Confucianism—fed into later world religions, Christianity and Islam among them. What societies dominated by Islam shared, allowing him to place them within a single framework of analysis, was what he took to be a moral project that synthesized and creatively extended Axial Age salvific visions.
Key to such Islamic moral teleologies was the distinctively Islamic formulation of prestigious sociality that this essay opened by explicating—namely, adab. Key to adab was a closed set of high-prestige, linguistic practices, the most prestigious of which was poetry. Already in the 800s and foundationally for a millennium, adab was the sort of rhetorically learned etiquette that gave evidence that a person was in command of the rational part of his (and sometimes her) soul. If, as Aristotle argued, the human animal was distinguished from the nonhuman animal by its rationality and if it gave interpersonal evidence of this rationality in controlled speech, then poetry, or “rhythmic and rhymed discourse,” was the most esteemed locus of this display of human moral self-fashioning.Footnote 20
In this sense, virtuous sociality was the necessary condition for literary language, a fact signaled in the Persian word sukhan, which means both “aestheticized speech” and “literature.” This also means that poetry and literary prose exerted their ethical effects only on readers (or listeners) who were prepared to receive them. Friendship, understood as a receptive hermeneutic disposition, was the ground for literary creation and ethical transformation. What made readers ethically receptive—what turned them into friends—was a variety of metaliterary discourses and practices. Premodern discourses on Arabic and Persian literature assumed and perpetuated a friendly habitus, a preconscious ethico-aesthetic disposition toward literature, but also toward experience in general, that was habitually made meaningful through literary metaphors just as literary metaphors themselves accrued new meaning through the life experiences in which they were invoked.
Any consideration of adab from a world-historical perspective would have to take seriously the hermeneutic postures that have formed this essay's topic and the models of friendship each assumes. But neither posture and the model of friendship each assumes resembles what Edward Said called “secular criticism” (1–30). That is, neither aims to read literature, as Said helped make normal for literary studies today, in ways attentive to its reproductions of power relations. The practice of charismatic allegory was as much an assertion of intergenerational guild power as the practice of humanist philology. Both entailed exclusively male authors, the privileging of Islam (even if not all such commentators were Muslims), and often inherited class and caste privileges by which the commentator accessed cumulative textual knowledge. In Said's terms, these were forms of affiliation that reproduced forms of familial and kin filiation. Even if the humanist philological mode was not as airtight in its interpretations as the Sufi allegorical one, its openness to accidental textual meaning was not motivated by an openness to marginal perspectives or minority consciousness.Footnote 21 (Ārzū, we must recall, treated such accidental linguistic effects as rule-making for philology.) In this sense, both forms of friendship and their respective interpretive styles, like the ideal and practice of civility they served, are valuable today to the study of Persianate literatures and, more widely, to world genealogies of virtue, as ethical and hermeneutic legacies of elite Persianate masculine corporate power.
