Hindu: A History

Abstract

This article provides a textured history of the multivalent term “hindu” over 2,500 years, with the goal of productively unsettling what we think we know. “Hindu” is a ubiquitous word in modern times, used by scholars and practitioners in dozens of languages to denote members of a religious tradition. But the religious meaning of “hindu” and its common use are quite new. Here I trace the layered history of “hindu,” part of an array of shifting identities in early and medieval India. In so doing, I draw upon an archive of primary sources—in Old Persian, New Persian, Sanskrit, Prakrit, Hindi, Marathi, Bengali, and more—that offers the kind of multilingual story needed to understand a term that has long cut across languages in South Asia. Also, I do not treat premodernity as a prelude but rather recognize it as the heart of this tale. So much of South Asian history—including over two thousand years of using the term “hindu”—has been misconstrued by those who focus only on British colonialism and later. We need a deeper consideration of South Asian pasts if we are to think more fruitfully about the terms and concepts that order our knowledge. Here, I offer one such contribution that marshals historical material on the multiform and fluid word “hindu” that can help us think more critically and precisely about this discursive category.

“You could say I’m Hindu. But…”

———Punjabi Dalit in Birmingham, 1990s¹

The Indian subcontinent is justly famous for its linguistic diversity, and yet a single word occupies my attention here: Hindu. The term “hindu” is ubiquitous today, rolling off tongues in dozens of languages in and beyond modern South Asia to denote a religious community. But what might seem semantically and conceptually unproblematic at first blush is, upon further investigation, complicated nearly beyond belief. After all, how do we define “hindu”? Scholars are often unsure, putting “Hindu” and “Hinduism” in scare quotes, to mark these broad-based identities as too fuzzy, too rigid, or anachronistic.² Practitioners, too, sometimes prefer to not call themselves “hindu,” instead seeking out alternatives, a point hinted at in the epigraph and to which I return at the end of this article. But try as some may, we cannot escape the term “hindu,” which permeates our speech and frames our thinking. In pursuit of better understanding the rich history of “hindu”—and usefully unsettling the term’s discursive power—I trace multiple strands of its elastic and wide-ranging meanings over 2,500 years.

Readers might object to the need and purchase of analyzing uses of “hindu.” In 2000, Donald Lopez stated that everyone is familiar with this bit of history: “the story of Indus to Sindhu to Hind to Hindu to Hinduism is well known by now” (832). Taking a different tact, Robert Frykenberg has argued: “‘Hinduism’ (if not the word ‘Hindu’) is a concept so soft and slippery, so opaque and vague, that its use all but brings critical analysis to a halt and intellectual discourse to the verge of paralysis (if not futility)” (1989: 87). I contend that both views are mistaken. Most scholars do not, in fact, know the contextual and semantic ranges of “hindu” in specific languages and historical moments, a story far richer than a basic etymological sketch. Moreover, “hindu” and “Hinduism” have never been empty signifiers, and I seek to recover their malleability in specific, not general, terms. In a way, my project is the opposite of a dictionary, where one collates commonly used words to distill knowledge (Lynch 2016: 11). Instead, I trace a single term to ground and sharpen our language and ideas through historical insights. In contrast to earlier scholars who have attempted akin projects, I draw upon a substantially more robust set of primary sources—in Old Persian, New Persian, Sanskrit, Prakrit, Hindi, Marathi, and more—that offers the kind of multilingual story needed to understand a term that cuts across languages in South Asia.³ Also, I do not treat premodernity as a prelude but rather recognize it as the heart of this tale. Much of South Asian history—including over 2,000 years of using the term “hindu”—has been misconstrued by focusing narrowly on the British colonial period and later. I offer a deeper consideration of South Asian pasts that can equip us to think more fruitfully about the identities that order our contemporary conceptual universe.

The word “hindu” has long existed alongside a set of related terms, including hind, hindī, hunūd (a plural), hindūstān, hindūī, hindavī, and hindūstānī. In pursuit of a manageable project, I do not trace this cluster of words, although I acknowledge slippage in their meanings. In particular, “hindu” often serves as a geographical and, later, linguistic marker. Both meanings are easier to grasp if we think of “hindu” within this assemblage of terms used to describe Indian cultures and societies. I include various spellings of “hindu,” including hinduka, hindush, hindua, himdu, hindura, and (in English-language sources) hindoo. Especially in the fourteenth century and later, “hindu” and associated terms are neither exclusively exonyms nor endonyms, a division that scholars have found fuzzy in many corners of the world (Woodman 2012). Critically, part of the story of “hindu” is precisely its multiformity and usage by discrete communities. Toward the end of the article, I discuss the advent of the English term “Hinduism” around 1800 and the neo-Sanskrit term “Hindutva” around 1920 because these abstractions have informed how many now demarcate and deploy “hindu.” Helpful to know at the outset is that use of “hindu” to denote a religious community occurs relatively late (fifteenth century CE at the earliest), and its use to describe or project followers of the broad-based religious tradition dubbed “Hinduism” today is far more recent. And so, recovering the history of “hindu” is, in part, an exercise of defamiliarization that, like the literary and artistic technique known by that name, can prompt us to reconsider afresh our shared vocabulary and attendant assumptions (Shklovsky 1998).

I do not attempt several things in this article. Most notably, I do not look for “Hindu” identities by other names, nor do I directly engage with the question of when Hinduism came to exist as a demarcatable religious tradition. Other scholars have addressed these subjects, most notably David Lorenzen in his seminal (and, for my project, inspirational) 1999 article, “Who Invented Hinduism?”⁴ The answer to that question depends on how you define “Hinduism.” A parallel body of scholarly literature has tried in recent decades to do just that, namely delineate Hinduism.⁵ This literature has proffered keen insights, but it has also left unexplored a promising venue of analysis, namely: how have others used “hindu” and “Hinduism” over time? In pursuing this question, I do not place undue emphasis on the origin of “hindu.” This is a subject of intense interest among Hindu nationalists today, stemming from V. D. Savarkar’s writings, in the anti-intellectual sense of privileging origins over usage.⁶ I reject that premise and the flattening of the past it enacts. Also, as I discuss in the postscript, Savarkar’s sense of “hindu” is a modern anomaly and so stands apart from the bulk of human thought and contestation over this term. Savarkar aside, readers may well find the malleable semantic range of “hindu” in South Asian history a contrast to its increasing rigidity in the twenty-first century. I think that may prove a valuable, rather than a merely disruptive, contrast. As we all contend with the weight of “hindu” as an organizing category, I argue there are insights to be gleaned from past usages, in meanings that linger today and in senses long forgotten.

Say My Name: The Non-Indian Debut of “Hindu”

Between the 500s BCE and 1000 CE, non-Indians used the term “hindu” to describe areas of northern India. Achaemenid inscriptions from the sixth century BCE constitute the first datable use of “hindu.” The multilingual Achaemenid Empire was based in Western Asia and stretched into the northwestern subcontinent (largely in modern-day Afghanistan and Pakistan). In inscriptions outside of India, the Achaemenids listed their landholdings, including an eastern part of their empire named hindush in Old Persian (sometimes with the -n elided), indu in Akkadian, and hinduš in Elamite.⁷ The Achaemenids inherited their geographical descriptor “hindu-” from the Indo-Iranian “sindhu” meaning “river.” Scholars remain uncertain about the origins of “sindhu,” which appears in the earliest known Indian text, the Rig Veda (ca. 1200 BCE), but has no clear Indo-Aryan etymology (Witzel 1999: 54). The term also appears in the Zoroastrian Avesta, such as in the expression hapta həṇdu (seven rivers) (Grenet 2015: 25–26). In its linguistic and cultural roots, “hindu” crisscrossed languages from the beginning.

After the Achaemenids, “hindu” was adopted in other languages in Western Asia to describe parts of India geographically. Greek authors beginning with Herodotus (fifth century BCE) used words derived from Old Persian “hindush” (spelled sans -h in Greek) for the subcontinent, especially northwestern parts thereof (Bivar 1988). Through Middle Persian, a version of the term entered New Persian in the late tenth century CE: hindū. ⁸ At this point, 1,500 years into the history of “hindu” as a geographical descriptor in a myriad of languages, we have no idea what or whether Indians thought about it. The term does not appear in early written Indian sources, whether in Sanskrit, Prakrit, or Tamil. It is unclear whether anyone on the subcontinent was aware of how they and their land were being described in Western Asia. But, if ancient Indians had known about the multilingual “hindu,” some may well have understood the impulse to define an Other. In a somewhat parallel activity, many Sanskrit intellectuals wrote during the first millennium CE about mlecchas or barbarians, who were defined by being located outside of āryavarta (land of the pure) and not participating in Brahminical Sanskrit culture (Thapar 1971).

Continuing an emphasis on defining the Other, the eleventh-century Arabic polymath al-Biruni made an early attempt to theorize a Hindu religion, although, importantly for my purposes here, not using the term “hindu.” Working under the support of Mahmud of Ghazna, Biruni penned an Arabic treatise on “the beliefs of the Indians,” widely known today as simply India (Lawrence 2000). Early in his India, Biruni proclaims unbridgeable differences between Hindus and Muslims, writing: “The people [of al-hind] entirely differ from us in every respect” (1887: 9). Biruni offers an early case of defining Hindus by contrast with Muslims, a theme that recurs—with varying definitions of the two communities—in later contexts and languages. But two caveats are important. One, Biruni’s work was a dead end. His India’s vision of a delimited Hindu tradition found no traction among premodern authors and, instead, “lay forgotten until it was resurrected in an even more radical form by European scholarship a century ago” (Ernst 2003: 177). Two, Biruni uses al-hind in Arabic. This is not merely a spelling quibble. Attention to precise vocabulary helps highlight that the history of theorizing Hindu traditions and the history of the multiplex term “hindu” are not coterminous, and yet we are liable to collapse the two due to modern preconceptions.

Poets, not philosophers, debuted “hindu” in Persian around the turn into the second millennium CE. In 1010, Firdawsi used hindūān in his epic poem Shāhnāmeh (Book of kings), for instance describing an “Indian ruler” (mihtar-i hindūān) (1997, 5: 552). Firdawsi put the phrase, appropriately given the term’s linguistic lineage, in the voice of Sikandar (Alexander of Macedon), whose Hellenistic kingdom encompassed parts of northwestern India.⁹ After Firdawsi, authors writing in various languages employed “hindu” to describe Indian kings. But more common in Persian poetry was the use of “hindu” to mean black, slave, or beloved, often woven into a contrast with the fair-skinned Turk.¹⁰ For example, Farrukhi (d. 1037) praised loving a congenial Hindu in contrast to the difficulty of winning over a stubborn Turk.¹¹ The Hindu-Turk contrast proved attractive to poets over the centuries, appearing in the famous couplet by Hafiz (d. 1390) that invokes a beloved’s beauty mark:

If that Shirazi Turk would grasp my heart in his hand,

I would give for his Hindu mole (khāl-i hindū) Bukhara and Samarkand.¹²

Persian poets also employed “hindu” as black when referring to the beloved’s tresses (zulf-i hindū) and the planet Saturn (hindū-yi falak; black doorkeeper of the sky). In one praise poem, Anvari (d. 1187) played with poetic meanings of “hindu” over numerous lines, speaking of an enchanting (afsūngar) Hindu beloved who both enflames and chars (1959, 1: 165–68).

The many poetic meanings of Persian “hindu”—including king, dark, slave, resident of India, and beloved—all appear in the works of Amir Khusraw, a prolific Indo-Persian intellectual who wrote within the Delhi Sultanate in the late thirteenth and early fourteenth centuries. For instance, Khusraw lamented:

Muslims have become sun-worshippers

Due to these saucy and innocent Hindus.

These pure Hindu boys

Have caused me to go to ruin and to drink.

Ensnared in their curly tresses,

Khusraw is like a dog with a collar.¹³

Khusraw also wove specificity into poetic representations of Hindus, perhaps because he lived in India (Sharma 2005: ch. 1). For instance, in his Hasht Bihisht (Eight paradises, 1302), he mentions the “languages of the hindus” (zabān-hā-yi hindūān).¹⁴ Elsewhere, he idealizes the practice of sati, where a widow immolates herself on her husband’s funeral pyre, exhorting: “Learn from the Hindu how to die of love! / It is not easy to enter the fire while alive!”¹⁵ Alyssa Gabbay has argued that, in addition to providing further depth to “hindu,” Khusraw reduced the contrast between Hindu and Turk at times, such as in this passage penned to honor a father-son truce that resolved a succession dispute:

Listen to this story: Jacob and Joseph

became one in this realm, without the torment of brothers.

O wind, tell the demons and fairies:

the two inheritors of Solomon’s realm have become one.

The Turks now do not travel to China

Hindustan is superior, since two emperors became one.

The duality has disappeared from Turk and Hindu

For Hindustan and Khurasan became one.¹⁶

These lines also bring in the older geographical angle of “hindu,” which proved resilient. Another example is the Hindu Kush mountains, first recorded by that name in the fourteenth century.¹⁷

Persian poetic invocations of “hindu” continued for centuries, into Ghalib’s poetry in the 1800s, but they stand apart from other premodern uses in one key respect: frequency. The term “hindu” found currency in the Persian poetic universe and so popped up, again and again, in poetry penned across Western and South Asia. In contrast, “hindu” was slow to gain traction in other languages. Even after Indians began using “hindu” beyond Persian in the fourteenth century, it remained infrequent. Often, “hindu” appears in a set of texts, is used a handful of times, and then evaporates for decades, even centuries. While leaving a light and seemingly disconnected series of footprints, Indians kept returning to “hindu” as an advantageous term for conceptualizing and positing at least vague contours of groups (usually) within the subcontinent in geographical, political, cultural, contrastive, religious, and other terms.

Hindu Goes Multilingual (Again) in the Fourteenth Century

In the fourteenth century, the term “hindu” went multilingual, appearing in Persian, Sanskrit, and Prakrit. The basic fact of multilingualism had precedent in Achaemenid inscriptions, but the geographical locations and communities had shifted. In the fourteenth century, Jains, Muslims, and people most moderns would anachronistically term Hindus employed this term in texts and inscriptions. Moreover, they did so from squarely within the subcontinent, in Gujarat, Delhi, the Deccan, and Tamil Nadu. These fourteenth-century cases all feature relatively sparse occurrences of “hindu,” meaning that the term appears in texts and inscriptions in a small number of uses or a single phrase. It seems that thinkers in many languages tried out “hindu,” but it did not dominate their writing. Additionally, fourteenth-century thinkers frequently demarcated “hindu” in two, often overlapping, ways: denoting a kind of king and defined by pairing with Muslims.

Jain authors used hindu, hinduka, and hindua in Sanskrit and Prakrit texts from the 1300s to establish a timeline of Indian rulers. The earliest datable example is the Vividhatīrthakalpa (Many places of pilgrimage, 1330–1333) of Jinaprabhasuri and Vidyatilaka, a text primarily about Jain pilgrimage sites. Both authors distinguished earlier hindua rulers from later miccha or anajja (Muslim) rulers. For instance, Vidyatilaka placed Hindu and Muslim kings in a chronology as protectors of Jain monks: “Just as when Hindus ruled (hiṃduarajje) in the Fourth Age, so too when non-Aryans rule (aṇajjarajje) in the Fifth Age, monks travel freely spreading the Jinas’ teachings.”¹⁸ Jinaprabha noted that “Muslim (miccha) kings will be powerful and Hindu (hiṃdua) kings weak.”¹⁹ Both authors accurately described contemporary political trends, especially the early fourteenth-century growth of the Indo-Muslim Delhi Sultanate. Critically, they judged Muslim and Hindu rule as comparably beneficial for Jain religious activities.

I remain uncertain how, exactly, Jain authors defined “hindu” rule and how they viewed it as distinct from “mleccha” rule. Christine Chojnacki has argued that hinduka and related terms mark a combination of religious, linguistic, and cultural affinities in early Jain sources (2011: 218). I see the evidence as strongest for linguistic difference. Some Jains, notably Jinaprabha, learned Persian at the Delhi Sultanate court (Vose 2013: 224–36). I can only speculate that this imperial connection is how Jinaprabha picked up the Persian “hindu.” We stand on firmer ground in seeing how Jain sources sometimes used “hindu” in contexts that hint at its Persian origin. For instance, Vidyatilaka named in succession “hindu kings” (hinduarāyāṇo) and “great lords” (mahāmalikkā), the latter adapting another Persian word (malik).²⁰ Other Jain texts invoke hinduka to describe a king or place in reported conversations with sultans or shahs.²¹ It seems that Jain authors associated “hindu” with Persianate contexts, even while none compared Hindus and Muslims so much as they more neutrally distinguished between them.

In the mid-fourteenth century, Vijayanagara rulers contrasted “hindu” and Muslim rulers in royal inscriptions lauding themselves as hindurāya-suratāla, “Sultan over Hindu kings.” The phrase first appeared in 1347 describing Marappa, a Vijayanagara founder: “Conqueror of the three kings, Lord of the eastern, western, and southern oceans, Vanquisher of kings who break their words, Destroyer of the pride of enemy kings, Lover of the courtesans of enemy kings, Sultan over Hindu kings, Victor over great enemy kings—Marapa is known by these titles.”²²

In subsequent decades, “Sultan over Hindu kings” appeared in numerous Vijayanagara inscriptions.²³ This striking title has been hailed as the first Hindu use of “hindu,” although the Vijayanagara rulers may have disagreed with that characterization (Talbot 1995: 700). After all, in fashioning themselves “Sultan over Hindu kings,” the Vijayanagara sovereigns claimed to be grand sultans who were superior to lowly Hindu rulers.²⁴ Moreover, as quoted above, “Hindu kings” is parallel to the conquered “enemy kings” (arirāya) in Marappa’s immediately surrounding titles. This slight against “hindu” rulers did not extend to Hindu religious activities, in which the Vijayanagara rulers regularly engaged. Moreover, a 1352 inscription that proclaims Bukka as “Prosperous great tributary, Punisher of enemy kings, Sultan over Hindu kings, Vanquisher of kings who break their word, Lord of the eastern and western oceans, and Auspicious hero” was placed within a Lakshminarayan temple.²⁵ Vijayanagara rulers used this title for the next 250 years, proclaiming political superiority to “hindu” kings.²⁶

“Hindu” also appeared in fourteenth-century Indo-Persian texts beyond poetry, most commonly in the coupling of “hindu” with “Muslim” or “Turk.” While Indo-Persian authors of this period were relatively consistent on this pairing, they differed wildly in its valence. For example, a Sufi text composed in Khuldabad, Maharashtra circa 1340 uses “hindus” without any religious connotation, when discussing battles involving Indians and Turkish soldiers.²⁷ In contrast, in his 1357 Tārīkh-i Fīrūz Shāhī, Barani—an Islamic hardliner who worked for the Delhi Sultanate—bemoaned that hindūs have forgotten God and fallen into debauchery and so too might Muslims (musalmān).²⁸ Barani was not alone in framing “hindu” as non-Muslim. Writing for the Bahmani court in the Deccan in 1350, Isami paired hindū and musalmān, elsewhere using hindī to mean Indian.²⁹ In addition to the shifting senses of “hindu,” one striking thing is that these uses offer no consistency—beyond the basic coupling—in how they frame the relationship between Hindus and Muslims. It points to an intriguing, persistent logic that defined “hindu,” by pairing with “Muslim” without anything further set about the two groups’ relationship.³⁰

A Rose by Any Other Name: “Hindu” and “Muslim” in the Fifteenth–Sixteenth Centuries

Between 1400 and 1600 CE, Indian authors invoked “hindu” in certain contexts to denote a religious community in Sanskrit and vernacular languages. Such usages were few and far between. They did not supplant older senses of “hindu” for people from a geographical region or, alternatively, a type of king. Even with these critical caveats, it is significant that a handful of thinkers—including Vidyapati, Kabir, Eknath, Guru Nanak, Jonaraja, and Shrivara—spoke about “hindus” in a religious sense circa 1400–1600 CE. Beyond that fact, the thinkers I survey here agreed on little. Notably, they varied on who they included within the “hindu” (or hinduka or hindura) community, with some referring only to upper-caste Hindus or Brahmins. For their substantial differences, all defined “hindu” by way of pairing. Most commonly, thinkers delineated Hindus alongside Muslims. In this sense, I agree with David Lorenzen, who observed more than twenty years ago: “Without the Muslim (or some other non-Hindu), Hindus can only be Vaishnavas, Saivas, Smartas or the like.”³¹ Most often, early modern thinkers appear to have been primarily concerned with delineating who “hindus” were not.

Vidyapati used “hindu” in a circa 1400–1410 Apabhramsha text to describe a community within the cosmopolitan north Indian city of Jaunpur. At the time, the Indo-Persian Sharqi dynasty ruled Jaunpur, which is described as follows in Vidyapati’s Kīrttilatā:³²

Hindus and Turks live together (hindū turuke milala vāsa)

One’s dhamma funny to the other

One calls the faithful to prayer. The other recites the Vedas

One butchers animals saying bismillah. The other butchers animals in sacrifices.

Some are called Ojhas, others Khojas

Some read astrological signs, others fast in Ramadan.

Some eat from copper plates, others from pottery.

Some practice namaz, others do puja. ³³

Here, Vidyapati advances, to my mind, two equally important claims. He equates Hindu and Muslim religious and cultural practices, positing comparable differences between their respective dhamme (Sanskrit dharma). Additionally, he names the two communities living together (milala vāsa) as a key feature of Jaunpur as a cosmopolitan metropolis.³⁴ Vidyapati may have felt free to include the newer vocabulary of “hindu,” because of his vernacular language choice of Apabhramsha as well as describing an Indo-Muslim political context.³⁵ In any case, both Sanskrit and vernacular thinkers used “hindu” in similar ways shortly after Vidyapati’s time.

In the fifteenth and sixteenth centuries, a mix of bhakti poets, Sikh leaders, and their biographers paired “hindu” with Turks or Muslims, often criticizing all. Kabir, a fifteenth-century bhakti poet claimed by both Hindu and Muslim communities, eschewed both identities, exhorting: “Kabir says: Worship the one Ram / Nobody is Hindu, nobody Turk” (hindū turk na kōī).³⁶ Kabir was echoed by the Sikh guru Arjun (d. 1606) who said, “We are neither Hindu nor Muslim” (nā ham hindū na musalmān).³⁷ Guru Nanak (1469–1539), the founder of Sikhhism, was remembered by sixteenth-century biographies to have invoked “Hindus and Turks” and “Hindus and Muslims” to reject both identities (Lorenzen 1995: 12). For instance, one account reports that Guru Nanak repeatedly proclaimed: “Nobody is Hindu, nobody Muslim” (nā koī hindū hai nā koī musalmān).³⁸ In these cases, thinkers paired Hindus with Turks or Muslims to point up the limits of engaging in religious pursuits within this binary. In later centuries, Indians from various traditions also played with positioning their spirituality as superior to that of both Hindus and Muslims, and we find such sentiments in Bullhe Shah, the Gorakh-bani, Todar Mal, et cetera.³⁹

Some thinkers compared Hindus and Turks to explore their differences and similarities. For instance, Eknath (1533–1599), a Brahmin bhakti poet who lived in Maharashtra, penned a dialogue between a Hindu and Turk (Hindu Turk Saṃvād).⁴⁰ The work begins with a clash:

The goal is one; the ways of worship are different,

Listen to the dialogue between these two!

The Turk calls the Hindu ‘Kafir!’

The Hindu answers, ‘I will be polluted—get away!’

A quarrel broke out between the two;

A great controversy began.⁴¹

As the debate unfolds, the two interlocuters heap insults upon one another as they take turns describing one another’s traditions. In so doing, they elaborate roughly (in Eknath’s view) parallel stories and practices, such as venerating icons (Hindu) and praying in the Kaaba’s direction (Muslim) (Zelliot 2003: 74). The poem ends with the Hindu and Turk embracing, after they realize that both traditions aim at higher truths. For thinking about what “hindu” meant for Eknath, two things seem critical. One, like Vidyapati, Eknath depicted Hindus and Turks as engaging in distinct but analogous religious activities, and so it was a difference without othering. Two, Eknath used “hindu” and “brahmin” interchangeably in the poem. Thus, while the terminology of “hindu” is eye-catching, the markers of this religious designation were, for Eknath, coterminous with Brahminical identity.

Other thinkers, too, used “hindu” to denote upper-caste Hindus or Brahmins specifically, such the Kashmiri authors of fifteenth- and sixteenth-century Sanskrit histories, each titled Rājataraṅgiṇī (River of kings). Four Kashmiri historians of this period used “hindu” or “hinduka”: Jonaraja (1459), Shrivara (1486), Shuka (1586), and Pseudo-Jonaraja (1575–1600, who interspersed 350-odd additional verses within Jonaraja’s history). All four authors employed the terms, relatively infrequently in their texts, to refer to high caste Hindus (dvija), often Brahmins specifically.⁴² All four also paired Hindus or Hindukas with Muslims, frequently when narrating local conflicts.⁴³ For instance, Shrivara referred to Brahminical customs (svācāraṃ hindukōcitam; ājanmahindukācārāś; hindukasamācāra) that were threatened by the behavior of Muslims (other times, he says Brahminical customs were supported by Muslims).⁴⁴ Shrivara also made other innovations. He coined the Sanskrit term mausula (from Persian musalmān) for Muslims and included significant information about Islamic cultural and religious practices.⁴⁵ These decisions attest to Shrivara’s substantial interest in thinking about religious communities in early modern Kashmir, although he arguably demonstrated more innovation regarding his descriptions of Muslims in Sanskrit rather than the well-worn idea of a Brahminical community.

Alongside the emphasis on high caste practices, “hindu” and “hinduka” also retained a geographical sense in the Kashmiri Rājataraṅgiṇīs. In an early verse, Shrivara refers to “Sindhu and Hinduvata lands.”⁴⁶ More interestingly, religious and geographical connotations came together in a story in Shuka’s Rājataraṅgiṇī (1586) where the Shahmirid king Muhammad Shah ordered the bones of deceased Hindukas taken to the Ganges to prevent mlecchas from disturbing them.⁴⁷ Invoking the Ganges’ cleansing power seems to define Hindukas, in part, by religious practices. At the same time, a marginal note glosses “hinduka” as “a resident of the subcontinent in vernacular” (hindukāḥ hindusthānīyāḥ hinde iti bhāṣayā).⁴⁸ Indeed, “hindu” never became particularly widespread in Sanskrit, although it cropped up in several further places through the eighteenth century.

Persianate, Jain, and “Hindu” Innovations in the Sixteenth–Eighteenth Centuries

Between the sixteenth and eighteenth centuries, uses of “hindu” still tended to be light, sometimes just a few mentions across a lengthy text. More generally, large swaths of intellectuals working in Sanskrit, Hindi, Bengali, and other languages did not employ this vocabulary. So far as I can surmise, “hindu” remained unattested altogether in some Indian languages, such as Tamil.⁴⁹ But the overall number of uses upticks in early modern Persian, Sanskrit, and north Indian vernacular sources. We find more robust uses of “hindu” to demarcate a religious community and, separately, an emerging sense of “hindu” to mean Rajputs, a specific category of Kshatriya kings in and around Rajasthan. Hindus (in modern terms) were among those who used the term in both senses, but they were not the sole drivers of either innovation. Also, no agreement emerged during this period on the contours of a “hindu” religious community. Here I present highlights of this sizable body of materials focusing on a Persian comparative treatise on religions, Jain Sanskrit works, and uses within Rajput, Maratha, and Gaudiya Vaishnava communities. Taken as a whole, “hindu” was endowed with further religious and political meanings between the sixteenth and eighteenth centuries.

Some early modern thinkers used “hindu” in Persian to denote a religious tradition, which could be compared with other religions, especially in Mubad’s Dabistān-i Maẕāhib (School of doctrines). Mubad, who followed the Iranian gnostic Azar Kayvan, penned the Dabistān-i Maẕāhib around 1650 near Hyderabad in southern India to ascertain the “truth of religion” (ḥaqīqat-i dīn) (Sheffield 2022: 960). Mubad covered a dozen religious traditions in the Dabistān, including a lengthy chapter on the practices and beliefs of hindūī (Hinduism) and hindūān (Hindus).⁵⁰ He often compares traditions, such as noting that Hindus, Jews, Magians, Christians, and Muslims all rely on scriptural law (sharīʻat).⁵¹ Mubad’s comparative structure constitutes a powerful way of constructing a Hindu tradition. In contrast, contemporary Sanskrit thinkers compared specific schools of thought, such as mīmāṃsā, sāṅkhya, nyāya, and so forth, in line with Sanskrit intellectual practices for centuries (Nicholson 2010). We might gather these traditions together today under the umbrella of Hindu philosophy, but Sanskrit intellectuals did not use that vocabulary. This backdrop highlights the important moment marked by the Dabistān-i Maẕāhib, which imagined a broad-based “hindu” religious tradition in detail and by comparison.

In the Dabistān-i Maẕāhib, Mubad defined the contours of a Hindu religious community differently from common ideas today, and this points to an easily overlooked mutability in early modern religious senses of “hindu.” For instance, Mubad identified a group “among the Hindus who consider themselves Sufi Muslims.”⁵² This view that Sufism is connected with Hinduism had legs and was repeated in various forms, including by early Orientalists, into the 1700s.⁵³ Mubad also described groups most would categorize as “Hindu” today, such as Nath yogis, as religion code-switchers who acted as Muslims among Muslims and as Hindus around Hindus (Ernst 2005: 41). These details from the Dabistān prompt us to raise questions about other, briefer uses of “hindu” in the seventeenth and eighteenth centuries. For instance, writing circa 1600 in Sanskrit, Devavimala compared Muslims (turuṣkajātīya) and “all Hindu classes” (samagrahinduvarga) in a passage that argues both worship comparable deities.⁵⁴ In Persian, Abdur Rahman Chishti (d. 1683) elaborated stories about Shiva, Vishnu, and other gods for a Hindu audience (hindūwān).⁵⁵ Writing in 1721 in Sanskrit, Lakshmipati posited that Kashi is to Hindus (hindūka) what Mecca is to Muslims (yavana).⁵⁶ It would be easy, if uncritical, to assume that these uses of “hindu”—in Persian and Sanskrit by a Jain, Muslim, and Hindu—are comparable to how most use the term today. But knowing that the Dabistān-i Maẕāhib defines “hindu” differently helps us to guard against that pitfall. Instead, we see that uses of “hindu” to refer to a religious community in early modern texts preceded set contours of that community.

Many who were interested in Hindu religious ideas elected to use alternative vocabulary, such as the Mughal prince Dara Shukoh (d. 1659). Dara supported Persian translations of elite Sanskrit texts, such as the Upanishads, and penned treatises proposing the similitude of Islam and elite Hindu traditions, most famously, Majma al-Bahrain (Confluence of two oceans). But “hindu” was not a key organizing category for Dara. As Supriya Gandhi has noted, Dara dismissed most Hindus as kāfir (infidel) and only engaged with a learned elite he dubbed muvaḥḥidān-i hind (monotheists of India).⁵⁷ Later Mughal thinkers described Dara as inclining toward the Hindu religion (dīn-i hindūān), but even so limited “hindu” to “Brahmins, Jogis, and Sanyasis.”⁵⁸ Dara Shukoh’s non-use of Hindu, as it were, helps to remind us that this vocabulary was still far from standard in the seventeenth century.

Early modern Jain Sanskrit intellectuals elaborated upon political senses of “hindu.” Most notably Padmasagara used “hindu” to mean Rajput several times in his Jagadgurukāvya (Poem on the teacher of the world, 1589) when discussing Mughal-Rajput struggles. In the sixteenth and seventeenth centuries, Rajput lineages contended with the growth of Mughal power, with most capitulating and all sharing with the Mughals significant cultural and political norms. In this set of political disputes, Padmasagara’s sympathies lay with the Mughals, who he said triumphed against “demonic Rajputs” (hindvāsura) and “demonic Rajput kings” (hindvāsurakṣmāpa).⁵⁹ In one section, Padmasagara narrated Uday Singh of Mewar’s reluctance to submit to the Mughal emperor Akbar. Padmasagara noted that many Rajput kings (hindunṛpa) wed their daughters to Akbar, which resulted in Akbar being the lover of both Rajput and Muslim women (hindumlecchasūtāḥ).⁶⁰ But, Padmasagara wrote, Mewar as the preeminent Rajput lineage (samastahindukalaśa) declined to follow suit, which led to military conflict.⁶¹

Given the emergence of “Rajput” as among the senses of “hindu” as a political category, it is perhaps unsurprising that the term proved attractive to Sanskrit and vernacular writers who eulogized Rajput lineages. Some combined “hindu” with other Persian terms. For instance, writing in 1585, Amrit Rai described the Kacchwaha ruler Bhagvantdas (r. 1573–1589) with two Persian titles: sadar-i sāhi vajjīr [vazīr] (foremost of imperial ministers) and hindu-panāh (shelter of Hindus [or Rajputs]).⁶² Amrit Rai perceived confluence between serving the Mughal crown and being an exemplary Rajput, for instance praising Man Singh and his son, Jagat Singh, as “Rajputs (hindū) [who] shone like a lamp” as they fought for Akbar.⁶³ Other poets who wrote for Rajput patrons positioned imperial service differently. Writing in the 1660s, Matiram praised the Bundi ruler Bhoj (r. 1585–1607/8) for “protecting Hindus’ pride” (hinduna kī rākhī sarama) and “rendering lame the foot of the shah’s authority (sāhi).”⁶⁴ Even within a delimited sense of “hindu” as Rajput, one might still ask whether there was any assumed religious content to the category? The answer varied by author, and here two thinkers—Narottam and Man Kavi—offer contrasting perspectives.

In a 1595 text written for the Kacchwahas, Narottam credited the Mughal emperor Akbar with instituting “Hindu rule,” defined by contrast with Turks and religious rites. He proclaimed that “Akbar, lord of Delhi, is praised across the four directions. / His is Hindu rule (haindū rāja). Who says it is Turk?”⁶⁵ Narottam next said that Akbar bathes in the Ganges, worships Hindu gods, reveres the Vedas and Puranas, and is himself an incarnation of Arjuna. Narottam reiterated his contrastive view toward the close of this passage, proclaiming: “Akbar loves Hindus (hinduna), he’s turned against the Turks.”⁶⁶ Here, there is religious content to “hindu” not dissimilar from how some might define the term today, and yet it is manifest in a Mughal (most would say Muslim) king. Writing close to a century later, Man Kavi, too, contrasted “hindu” and Muslim, but by injecting new blood into an older framework of “Kshatriya dharma.” Writing in Pingal (a Rajasthani form of Braj Bhasha) in 1680, Man Kavi praised the Mewar ruler Rana Raj Singh (1652–1680) as “lord of Rajputs” (hindūpati) and “maintainer of Kshatriya dharma” (Talbot 2018: 472). Later, Man Kavi envisioned Rajput domination over the Muslim Mughals, while defining “hindu” as upper castes alone:

I spread the superior Veda and will preserve on earth the Puranas,

The qazi’s books and all the Qurans, I reduce all these to ashes,

I will grind down the Chagatai and establish my own garrison in Delhi,

I maintain Hindu customs (hiṃdū rīti) and uproot the demonic ways,

I will raise up the best holy temples and tear down the mleccha sites,

I will protect all the Rathors, the angry Rana Raj said.⁶⁷

This sort of rhetoric invoked a caste-ordered past where Kshatriya kings were supposed to uphold upper-caste privileges. But Man Kavi specifically contrasted Kshatriya and Mughal kingships, complete with a promise to suppress Muslim practices. Even as more authors grew interested in thinking about “hindu” kingship, they found no agreement on its contours.

The Maratha courts of Shivaji and his descendants constitute a second group of Hindus (in modern terms) that used “hindu” in this period, but they did so sparingly. I find this unsurprising given Shivaji’s anxieties about being born in the Shudra varna, the lowest of the four classes, and his consequent concern with projecting himself as a Kshatriya Rajput.⁶⁸ For instance, in his 1675 Sanskrit Sūryavaṃśa (Dynasty of the sun; better known as Śivabhārata), Paramananda described Shivaji as kṣatriya and bāhuja (Kshatriya).⁶⁹ Likewise, Paramananda narrated how Shivaji maligned his enemy Afzal Khan as “hell-bent on obstructing the path of caste dharma (varṇadharma).” A similar concern with preserving caste hierarchy surfaced in other Maratha works, such as Keshava’s 1690 Rājārāmacarita (1931).⁷⁰ Being depicted as a “hindu” king offered space for novelty in this period but not the longevity and traditional weight of being a Kshatriya ruler for which Shivaji longed.

Some early moderns used “hindu” in vernaculars to refer to a religious community, although its contours and exclusions were inconsistent, with one another and with the term’s most common sense today. For example, five Gaudiya Vaishnava hagiographies written in Bengali between the early sixteenth and the late eighteenth centuries used “hindu” and “hindura,” mainly in contrast to Muslims (using the vocabulary yavana, mleccha, pathan, and turk).⁷¹ Most passages identified a mix of religious and cultural norms. For instance, the texts refer to the “Hindu god” (hindura īśvara) and “Hindu treatise” (hindu-śāstre), on the one hand, and to “hindu clothes” (hindu-beśa), on the other (O’Connell 1973: 341). One of the more intriguing features of “hindu” in these texts is that it is, at least once, contrasted with Vaishnava identity. In Krishnadasa Kaviraja’s Caitanyacaritāmṛta (1612–1615), one group argued that singing aloud God’s names is against “hindura dharma,” and Vaishnavas defended the practice (ibid.: 342). This passage follows the general trend of defining Hindus by contrast, even while the other group is one that, today, nearly everyone would include within the Hindu tradition. A 1785 usage from the other side of India offers a harsher contrast, wherein the Rathor state of Jodhpur distinguished hinduvāṃ from achhep, Marwari for “untouchable.” As Divya Cherian points out, the Jodhpur crown defined upper-caste Hindus by contrast, specifying that definitionally non-Hindu achhep communities included Muslims, leather workers, agricultural laborers, hunters, and human scavengers (2022).⁷² This was neither the first nor the last time that those who claimed the identity of “hindu” cited caste as a key consideration in defining their tradition.

New Outsiders and New Terms: Hindoo, Hindooized, and Hinduism

During the nineteenth century, “hindu” became widespread in many Indian languages and English. The broad shift from light, if recurrent, premodern uses of “hindu” to this term being an organizing category of identity featured specific European and elite Indian communities who encountered one another within the colonial context. During this process, several key things happened. European English-speakers coined numerous new terms, including Gentoo, Hindooized (often with that spelling), and, most critically, the abstract noun Hinduism. Through cross-cultural conversations, these terms had significant ramifications for the vocabulary and, arguably, the self-understanding of specific Indian communities. In both European and Indian uses, “hindu” (or “Hindoo”) was imbued with a stronger religious connotation in this period, although it remained broad enough to encompass other senses too (chiefly geographic, political, and linguistic) and to enfold groups most would categorize as distinct religions today (e.g., Sikhs and Jains). In addition, access to certain languages—especially Sanskrit and Hindi—emerged strongly as part of what “hindu” could denote. Even as “hindu” remained a fluid term in the colonial period, it went viral, becoming an indispensable part of the vocabulary of Europeans and Indians alike.

Europeans adapted “hindu”—often with the English spelling “Hindoo”—to refer to residents of India, a religious community, and an elite tradition of learning. In a notably early usage, Edward Terry, a British East India Company chaplain, used “Hindoo” in 1616 to gloss “Inhabitants of Indostan,” also known as “Gentiles.”⁷³ This usage persisted for centuries. Although, from the eighteenth century onward, Europeans increasingly used “Hindoo” when discussing Brahmins and their Sanskrit texts. For example, writing in 1770, Alexander Dow referred to “Hindoo faith” and “Hindoo religion,” sometimes citing Brahmins and the Vedas.⁷⁴ Others spoke specifically of “Hindoo Learning” or, continuing a legacy begun with “Gentoo,” “Hindoo law” and “Hindu law.”⁷⁵ In the late eighteenth century, the British patronized traditional forms of Indian learning, such as founding Benares Sanskrit College in 1791 as a “Hindoo College,” in the words of Jonathan Duncan, the British Resident at Benares.⁷⁶

Some eighteenth-century Europeans preferred the term “Gentoo,” derived from Latin gentile through Portuguese gentio, when discussing elite Brahminical traditions. This slotted Hindu traditions into a common European Christian conception of a fourfold division of world religions: Christianity, Judaism, Islam, and Gentiles or Pagans (Masuzawa 2005: 46–69). For instance, in the 1750s, Jonathan Holwell lamented losing “Gentoo manuscripts,” including a “Gentoo Shastah.”⁷⁷ Two decades later, Nathaniel Halhed published A Code of Gentoo Laws, a collation of Brahminical Sanskrit lawbooks. In his preface, Halhed used as synonyms “Hindoo” and “the Brahminical Religion” (1776: xxii–xiii). In a 1779 letter, he similarly used all three terms—Brahmin, Gentoo, and Hindoo—including a section defending his use of “Brahmin” (spelled Brachman) because it is found “among all the Laws of the Gentoos” (5b; original underlining). Soon European thinkers abandoned “Gentoo,” settling instead on describing aspects of Indian cultures—especially elite, upper caste, Sanskrit traditions—as “Hindoo” or “Brahmin,” along with their respective abstractions: Hinduism and Brahminism. I return to the importance of abstracting from “hindu” to “Hinduism” below, although it is worth noting that this too finds precedent in Holwell, who used (and perhaps coined) “Gentooism.”⁷⁸

“Hindoo” could also be a porous identity in colonial-era European uses, and Muslims, Christians, and Europeans who adopted certain manners and customs were sometimes labeled “Hindooized” or “half-Hindooized.” At times, the issue seemed to be Europeans who went native. For example, George Campbell of the Bengal Civil Service wrote in 1852 about Europeans (his term) “who have become enthusiasts in admiration of the natives, and partially ‘Hindooized’” (295). Other times, European authors expressed anxiety about Indian-seeming Christian practices, such as the travel writer Anna Harriette Leonowens who wrote of her experience in India (1897: 56): “One finds everywhere in India not only Hindooized Mohammedans, but Hindooized Christians. Their priests are natives of the country, under the jurisdiction of the archbishop of Goa, who is a sort of Indian pope. Their worship is so much more pagan than Christian that when in a Roman Catholic church in any part of India one finds it difficult to believe that it is not the worship of Khrishna or Brahm.” According to other Europeans, Muslims were most at risk of being “Hindooized,” both by practicing “idolatry” and observing caste.⁷⁹ Taking “Hindooize” in the other direction, M. A. Sherring, a missionary in British India, wrote in 1868 that “education de-Hinduizes the Hindu” specifically because it “breaks down idolatry, and inspires him with a distaste for it, and a latent desire to be free from it” (350). In these usages, the term “Hindoo” appears defined by social practices and so could be an identity adopted, or slipped into, by anyone present on the subcontinent.

Other Europeans used “hindu” as an exclusive religious category, perhaps most overtly in the invention of the term “Hinduism.” So far as scholars have discerned, Charles Grant, an evangelical Baptist convert, coined the term “Hindooism,” with that spelling, in a 1787 letter written from Calcutta to a friend in England.⁸⁰ Grant later used “Hindooism” numerous times in books and letters, as did others associated with the Danish Baptist mission at Serampore.⁸¹ Grant more frequently used the established expression “Hindoo religion,” and so it is unclear what, if any, innovative content he ascribed to “Hindooism.”⁸² What is clear is that Grant and the Baptists at Serampore sought Hindu converts, and so the word “Hindooism” was birthed as part of how some Europeans conceptualized the missionary field and their desire to Christianize the world.⁸³ In this effort, Hinduism was often contrasted to Christianity, a comparison that pointed to perceived doctrinal clashes and, in some discourses, supplanted the older Hindu-Muslim pairing. Critically, “Hindu” as a category was not porous for these men, with William Ward even lamenting that missionary efforts were unlikely to yield significant conversions since “we cannot become Hindoos to win them [Hindoos]” (Stanley 1990: 160).

The term “Hinduism” gained wider currency throughout the first half of the nineteenth century, including among Hindus. In 1816, Rammohan Roy, a Hindu reformer and Bengali, was perhaps the first Hindu to use “Hindooism.” Roy probably picked-up the term from Serampore missionaries, one of whom he met in 1815 (Oddie 2010: 45). In this earliest known self-referential use of “Hindooism,” Roy defined the tradition by social practices associated with caste: “For the chief part of the theory and practice of Hindooism, I am sorry to say, is made to consist in the adoption of a peculiar mode of diet; the least aberration from which, (even though the conduct of the offender may in other respects be pure and blameless) is not only visited with the severest censure, but actually punished by, exclusion from the society of his family and friends. In a word, he is doomed to undergo what is commonly called loss of cast.”⁸⁴ Roy was part of a larger group of Hindu reformers who interacted with, responded to, and adapted Christian criticisms of Hindu practices and beliefs (Thapar 2019). And so, Roy did not adopt a well-worn term so much as he entered debates about how to define (or, for him, redefine) this tradition, including by conceptualizing it as unified or at least demarcatable. The idea of discussing the Hindu tradition, both singularly and more abstractly, is also attested in Marathi and Hindi sources of this period, using the phrases “hindu-dharma” and “hindumat.”⁸⁵ Some of these texts explicitly sought to defend “hindu-dharma” against Christian criticisms.⁸⁶

As “Hinduism” and “Hindu” became increasingly common terms throughout the nineteenth century, agreement on their definitions was slow to emerge. John Crawfurd used “Hinduism” in an 1820 scholarly article on Hindu practices in Bali, a clear indication that the tradition was not geographically confined to the subcontinent.⁸⁷ But Webster’s 1828 dictionary defined “Hin’doo” as “an aboriginal of Hindoostan, or Hindostan.”⁸⁸ A 1829 book on Bengal by Henry Henderson distinguished Marathas and Hindus (“Mahrattah” and “Hindoo”) (312). Whereas a Hindu tract society, founded in Madras in 1887, argued that Hindus should not subdivide themselves as “Saivites, Vaishnavites, Advaitins” and so forth and instead should unite to oppose Christianity.⁸⁹ At times, these variations in meaning seem similar to the longstanding flexibility of “hindu.” But other times, such as in fierce dialogues comparing Hinduism and Christianity, individuals outlined clear stakes in advancing specific contours to the tradition increasingly called “Hinduism.” For instance, writing in 1817, Roy argued that “the doctrines of the unity of God are real Hindooism,” as opposed to “the superstitious practices which deform the Hindoo religion” (Roy 1817). Scholars, too, weighed-in on these debates. Especially notable is that Monier-Williams, Boden Professor of Sanskrit at Oxford, published an 1878 book in which he separated “Brāhmaṇism” from “Hindūism” (his spellings), advancing a definition for the latter influenced by Hindu reformers.

As “Hinduism” entered into common usage to denote a religious tradition (albeit with contested contours), “hindu” continued to have geographical and linguistic connotations, especially within the context of emerging nationalism. For instance, the Hindi slogan “hindī, hindū, hindustān” was coined by Pratap Narayan Mishra in the late nineteenth century with these lines:

If you truly desire your own welfare

then keep chanting this mantra with one tongue, hindī, hindū, hindustān

whether it attracts or repels the world, brings respect or affront

don’t leave off chanting with one voice, hindī, hindū, hindustān

those who don’t know their own identity are like the living dead

so sing loud this grand mantra, hindī, hindū, hindustān

wise are those who don’t discard their own language, food and clothing

in all its proof of good fortune, hindī, hindū, hindustān.⁹⁰

This call for identity interweaves language, religion, and geography in new ways. Whereas earlier uses of “hindu” had sometimes involved Sanskrit, here the claim is that Hindi is, to quote Grierson in 1889, “the lingua franca of Hindūs.”⁹¹ Another aspect of Mishra’s innovation is exhorting people to strongly proclaim a layered Hindi-Hindu-Hindustan identity, a point in sharp contrast to the infrequent and light uses of “hindu” in earlier Indian texts. Indeed, that “hindu” was increasingly a contested identity, rather than a fluid and flexible one, also comes out in the writings of Mishra’s associate, the Hindi author and lay religious leader Harishchandra of Banaras (1850–1885). Harishchandra argued that “whoever lived in Hindustān, whatever the colour or the jāti, was a Hindu” (Dalmia 1997: 26). This expansive sense of “hindu”—including to cover people belonging to other religious traditions—persists today and is even enshrined in Indian law. Article 25 of the Indian constitution specifies that the category of “Hindu” legally includes “persons professing the Sikh, Jaina or Buddhist religion.”

Still, for others in British India, “hindu” remained a fuzzy, even irrelevant, identity. This comes out in one government exercise that does not permit fuzziness: the census. British officials used “hindu” even in early census exercises, such as in the 1823 Census of Benares when Prinsep noted that the city proper was “almost exclusively Hindu.”⁹² But, as the British sometimes objected, the people that they counted as “Hindu” did not always agree. For instance, in 1921, the Census Commissioner lamented: “The chief hindrance to the obtaining of accurate returns is the fact that the terms used to classify the religions are unfamiliar to the people of the country.… No Indian is familiar with the term Hindu as applied to his religion. If asked what his religion is, he usually replies with the name of the sect (e.g., Saivite), to which he belongs.”⁹³ Here, 2,400 years into the history of the term “hindu,” many Hindus still did not embrace this label, an intriguing, historically recurrent part of the journey of this polyphonic term.

Postscript: Ethnonationalism, Race, and Other Modern Ideas

In usages over the last century, “hindu” has most often referred to a religious community but not always. The enduring porousness of “hindu” recalls the term’s multivalent past, but, critically, some of its modern incarnations are decidedly restrictive in their meaning. In the first half of the twentieth century, “hindu” served as a racial category in United States discourse. At the time, few Indians—a total of 4,901 in 1920, largely from Punjab—lived in America (Chakravorty, Kapur, and Singh 2017: 6). The United States had some exposure to Hinduism as a religion, especially through Vivekananda, a “Hindoo monk of India” who lectured across the United States in the 1890s (Kaplish 2019). But, Americans were seemingly more interested in thinking about Indians—in contrast to their abstracted religious traditions—in racial terms. For instance, in 1907, a nativist mob expelled several hundred Punjabi laborers—mostly Sikh and about 10 percent Muslim—from Bellingham, Washington. News reports described the displaced laborers as “East Indians” and “Hindu hordes” (SAADA 2021: 29). On the 1920, 1930, and 1940 United States censuses, “Hindu” was a racial category used for all South Asians.⁹⁴ A 1923 United States Supreme Court case centered on the question of whether “Hindu” and “white” could overlap as racial categories (the court decided they could not).⁹⁵ These uses of “hindu” harkened back to earlier geographic delimitations in some ways, but the racialization of “hindu” identity was new and soon forgotten in all but one context.

In the 1920s, V. D. Savarkar popularized the term “Hindutva,” a combination of the Persian hindū with the abstract Sanskrit suffix -tva, to describe the political ideology also known as Hindu nationalism. Linguistically, “Hindutva” is a calque of “Hinduism,” but Savarkar was an atheist who mocked Hindu religious practices such as cow veneration and so sought to substantially redefine “hindu.”⁹⁶ In his most famous book, Hindutva: Who Is a Hindu?, Savarkar argued that a “hindu” possesses three characteristics: being a native resident of India, having a specific racial lineage, and participating in Sanskrit (really, neo-Sanskrit) culture.⁹⁷ This definition of “hindu” took little from earlier Indian thought and was instead heavily indebted to ethnonationalist movements in early twentieth-century Europe, including Nazi ideas about the Fatherland and Aryan racial superiority.⁹⁸ Post-Savarkar, some Hindutva groups have used more religious language, but often in imitation of Protestant ideas (Waghorne 2004: 18–19). As Jack Hawley once described “Hindutva” in a riff on the penchant of the Rashtriya Swayamsevak Sangh, a paramilitary Hindutva group, for khaki shorts in imitation of British colonialists: “The raw material and the idea of a half-size pant may have been Indian, but the cut, definition, and standard ritual usage came from Europe” (1991: 22). In the robust intellectual genealogy of the term “hindu,” the advent of “Hindutva” constitutes a historical break where Indian thinkers attempted to fetter a fluid term and weaponize it to exclude and oppress. It remains to be seen whether Hindutva ideologues will succeed in redefining “hindu” in broader usage, but I would not hold your breath. The term “hindu” has long proved difficult to pin down to a single meaning.

Today, “Hindu” and “Hinduism” are predominantly religious categories, used widely by practitioners and scholars alike, except that both groups exhibit some discomfort with these labels. Scholars sometimes qualify “Hinduism” for specificity, speaking of syndicate Hinduism, bhakti Hinduism, tribal Hinduism, temple Hinduism, village Hinduism, and so forth (Frykenberg 1989: 87, 90). Practitioners sometimes eschew the label of “Hinduism,” saying, for example, that they follow sanātana dharma. Followers of “sanatana dharma” partake of a longer legacy—which has recurred throughout this article—of Hindus who prefer to not call themselves Hindu. That said, their self-description of “sanatana dharma” dates to the nineteenth-century Hindu reform movements. In this sense, it is useful to consider “sanatana dharma” alongside other terms coined in the colonial era, such as “Hinduism” and “Brahminism.”⁹⁹ No doubt this contextualization will surprise some followers of “sanatana dharma” who, after all, use that neo-Sanskrit phrase precisely to project their tradition as “eternal” and “universal” (Iskcon n.d.). But the gulf between community projection and historical reality illustrates part of what I have tried to uncover throughout this multilingual history of “hindu,” namely usefully destabilizing what we thought we knew by introducing a multisource historical narrative. Going forward, I imagine we will all continue to use the terms “hindu” and “Hinduism,” although hopefully with a more critical appreciation for our options and agency in defining these malleable categories.