Savarkar’s Hindutva of 1923 is the most argued-over short book in Indian political thought, and much of the argument proceeds without agreement on what the book says. The pamphlet that V.D. Savarkar composed in a Ratnagiri prison cell and published in 1923 under the borrowed name A Maratha ran to eighty-eight pages, and those pages contain a definition before they contain a programme. That order matters. This article defends a namable claim: Savarkar’s 1923 framework was territorial, ancestral and civilisational rather than theological; it sorted human beings by land, lineage and the geography of sacred memory, and it deliberately refused to sort them by belief. The conceptual-architectural reading, which attends to how the definition is built, is analytically prior to the political-ideological reading, which attends to what later actors did with it. Only the architectural reading makes the rest of the record intelligible: the author’s atheism, his assault on caste prohibitions in the Ratnagiri years, his contempt for scriptural literalism, and his insistence that a Charvaka materialist and a temple priest could stand inside the same identity. Those commitments puzzle readers who meet the text through its political afterlife. They follow with mechanical consistency from the definition’s design.

The disagreement the article adjudicates is named on all sides. Vinayak Chaturvedi, in Hindutva and Violence: V.D. Savarkar and the Politics of History (2022), reads the 1923 text as the foundation document of a political project whose organising figure is the enemy, and he traces the consequences through a century of mobilisation. Janaki Bakhle, in Savarkar and the Making of Hindutva (2024), reconstructs the making of the concept across Savarkar’s Marathi and English writings and treats the literary and affective labour of definition as the real event of 1923. Christophe Jaffrelot, whose Hindu Nationalism: A Reader (2007) placed extracts of the text before students of the subject, situates it in a genealogy of Hindu nationalism shaped by stigma, fear of the other and emulation of the other’s imagined unity. The adjudication offered here does not ask readers to set the political record aside. It asks them to read the architecture first, because the political activities of the Hindu Mahasabha years operated inside the framework the definition built, and because a framework misread at the level of its criteria will be misread at every later level too. The approach continues the method set by the tripartite-disentanglement framework, which separates the Hindutva variants before judging any one of them; Savarkar’s 1923 text is the first variant, and its internal design is the evidence.

The date of the book fixes its position in the genealogy this pillar reconstructs, and the position is earlier than casual readers assume. The text of 1923 precedes the founding of the Rashtriya Swayamsevak Sangh at Nagpur in 1925, precedes M.S. Golwalkar’s We or Our Nationhood Defined of 1939 by sixteen years, and precedes every organisational form through which the ideas it contains later travelled. Genealogy is not causation, and this article does not argue that the Sangh is the book’s creature; K.B. Hedgewar built his organisation from his own experience of Congress volunteer corps and revolutionary societies, and the debt ran through persons, Kelkar of Nagpur among them, rather than through a paper chain of command. The priority of the text is nevertheless a fact with consequences for method. A reader who meets Hindutva first in its later organisational forms imports the organisation’s questions into the pamphlet: questions of discipline, of the nation as an organic body, of electoral mobilisation. The pamphlet cannot answer those questions, because it was written before they were posed in those forms, by a prisoner whose problem was classification under conditions of confiscation. Restoring the text to 1923, ahead of its descendants, is the first act of the architectural reading, and the disentanglement of the variants that the pillar hub performs begins with the restoration.

Savarkar and Hindutva: The Conditions of Writing in Ratnagiri

The conditions under which Essentials of Hindutva was written are part of its meaning, because the book is the product of a revolutionary removed from every instrument of politics except definition. Vinayak Damodar Savarkar was born on 28 May 1883 at Bhagur in Nashik district, was arrested in London in 1910 on charges that included abetment of murder and waging war against the King-Emperor, escaped briefly from the SS Morea at Marseilles in July 1910, and was sentenced in Bombay to two terms of transportation for life. He reached the Cellular Jail at Port Blair in July 1911 and spent a decade there, composing poetry, teaching fellow prisoners to read and, by his own later account, working out the leading arguments of the definition in his head because pen and paper were denied him. The front matter of the 1942 edition of the book records that released prisoners carried the gist of the treatise and the definition itself back to India in memory between 1917 and 1919, so the text had an oral prehistory inside the prison system before it had a printed one.

The commutation that followed his clemency petitions brought him to the Indian mainland on 2 May 1921. After a week at Alipore jail and a passage through Bombay prisons, the Government of India lodged him in the district jail at Ratnagiri, where he remained for under two years. There, with writing materials obtained at intervals through fellow political prisoners who enjoyed facilities denied to a revolutionary convict, he wrote the treatise out in full. In May 1923 the manuscript was smuggled from his cell to Vishwanathrao Kelkar, an advocate at Nagpur and a comrade from the Abhinav Bharat years, and Kelkar published the eighty-eight page English pamphlet in Nagpur in 1923 under the pseudonym A Maratha. The author’s name could not appear: Savarkar was a serving prisoner, and attribution would have invited suppression of the book and punishment of the author. He left prison only in January 1924, on conditions that confined him to Ratnagiri district and barred him from political activity; the internment lasted until May 1937.

A definition produced under those constraints bears the constraints in its form. Deprived of a party, a press and a public platform for sixteen years after 1921, Savarkar built the one artefact a prisoner could build: a classification. The book argues almost nothing about organisation, elections or tactics. It asks a prior question, who is a Hindu, and it answers that question with three criteria arranged like the clauses of a statute. Readers who approach the pamphlet expecting a political programme find a taxonomy instead, and the mismatch between expectation and artefact explains much of the secondary literature’s confusion. The political-ideological reading treats the taxonomy as a programme in disguise. The architectural reading takes the taxonomy at face value and asks what the taxonomy, exactly as designed, commits its holder to accept. That question has precise answers, and the text supplies them chapter by chapter.

Where was Savarkar held when he composed the manuscript?

Savarkar composed the manuscript in the district jail at Ratnagiri on the Konkan coast of Maharashtra, where the Bombay authorities lodged him after his 1921 repatriation from the Cellular Jail. He wrote with materials obtained intermittently through fellow political prisoners, and the finished text left the prison by smuggling in May 1923.

The road to that cell explains the prisoner’s authority to define, because Savarkar arrived at Ratnagiri with a revolutionary record that the definition’s later readers sometimes forget and sometimes inflate. At India House in London, the hostel for Indian students that Shyamji Krishna Varma maintained at Highgate, the young law student from Fergusson College had organised the Free India Society, celebrated the anniversary of 1857 as a national festival, and written the Marathi life of Joseph Mazzini that carried the Italian’s doctrine of national resurrection to Maharashtra in disguise. His The Indian War of Independence of 1857, written in London and printed in 1909, recast the rising as a national war and was proscribed by the Government of India before a copy could lawfully circulate. The arrest in 1910, the leap from the SS Morea into the Marseilles harbour in July of that year, the recapture on French soil and the arbitration at The Hague that returned him to British custody made him, at twenty-seven, the most famous Indian prisoner in the empire. The Nasik conspiracy proceedings and the trial for the murder of the collector A.M.T. Jackson, shot at Nashik in December 1909 by Anant Kanhere with a pistol traced to Savarkar’s circle, produced the two life sentences he carried to Port Blair.

The Cellular Jail decade deserves its weight in the argument, because the definition’s portability was learned there. The regime at Port Blair ground prisoners at the oil mill, held them in solitary cells and treated the literate revolutionary as an object lesson. Savarkar’s response, recorded in his autobiography Mazi Janmathep and corroborated in outline by the jail’s own historiography, was pedagogical: he taught illiterate prisoners to read, organised a clandestine library from smuggled books, and composed verse that inmates memorised as a circulating archive. A man who watches every material instrument of politics confiscated learns to value the instrument that cannot be confiscated, and the 1923 definition has that valuation built into its form. It requires no temple, no press, no party and no priest. It can be carried in the head across an ocean in irons, as its author’s own prisoners carried its gist to India between 1917 and 1919. The political-ideological reading, attentive to the definition’s later institutional career, underweights this origin. The architectural reading begins with it, because the origin explains the design’s most distinctive feature: the definition’s complete indifference to the apparatus of religion, which the prisoner did not control, and its complete reliance on land, descent and memory, which no jailer could impound.

The text also arrived with a deliberate modesty of dress. The 1923 edition presented itself as an inquiry into essentials, an attempt to state the minimum conditions of an identity rather than its maximum programme, and the 1928 reprint gave the inquiry its enduring interrogative title, Hindutva: Who Is a Hindu? The question form is the architecture in miniature. A creed announces what its adherents must believe; a definition announces what its subjects must satisfy. Savarkar chose the second form, and the choice governed everything the framework later permitted and forbade, including positions that scandalised the orthodox of his own camp. The sections that follow take the definition apart piece by piece, then test the two great readings of the book against the evidence of the pieces, then examine the biographical commitments, atheism and anti-caste practice chief among them, that the pieces predict.

The book’s publication history after 1923 completes the conditions-of-writing argument, because the text kept being reissued at the moments when its author needed it to work. The 1928 reprint, published when the internment had five years still to run, gave the book its question for a title and carried the argument to a wider public under the author’s own name. The edition of the early 1940s, prepared as the Mahasabha presidency made the author a national figure, clothed the text in the front matter the historian relies upon: the account of the Andaman gestation, of the oral transmission of the definition by released prisoners between 1917 and 1919, and of the tributes the first edition had gathered from Lala Lajpat Rai, Madan Mohan Malaviya and Swami Shraddhanand. Translations into Hindi and the Indian languages followed across the century, and the Mahasabha’s adoption of the culminating sentence as its official definition made the pamphlet a constitutional document for a movement, quoted in resolutions the way statutes are quoted in pleadings. Each reissue confirms the design thesis from a fresh angle. A programme dates, and must be revised or discarded; a definition is reprinted, because its holders expect the classification to go on sorting long after the circumstances of its composition have passed into history.

The Definition Among Its Contemporaries: Sangathan, Shuddhi and the Khilafat Years

No text of 1923 wrote itself in a vacuum, and the architectural reading gains rather than loses by placing the definition among the competing answers to the same decade’s question. The years of the book’s gestation were the years of the Khilafat movement, when the Ali brothers and Mahatma Gandhi bound the Congress to the cause of the Ottoman caliphate, and of the Moplah rising of August 1921 in Malabar, whose violence against Hindu neighbours entered the Hindu press as a parable of disorganisation. The Hindu Mahasabha, founded at Haridwar in 1915 and reinvigorated at its Benares session of 1923 under Madan Mohan Malaviya’s presidentship, supplied the organisational shell of a response whose keywords were sangathan, consolidation, and shuddhi, the readmission of converts. Swami Shraddhanand, the Arya Samaj sanyasi who led the shuddhi campaigns until an assassin killed him in December 1926, preached the consolidation of Hindus as a programme of schools, gymnasiums and reconversion. Against that field, the distinctiveness of Savarkar’s contribution stands out in relief. The sangathan movement had slogans, grievances and campaigns; it lacked a criterion. Who exactly was to be consolidated, and by what test would the consolidators know their own? Shraddhanand’s movement answered by practice, consolidating whomever the practice reached. The prisoner at Ratnagiri answered by definition, and the definition’s exclusions and inclusions, so often read as the movement’s prejudices transcribed, are in the text the output of a test the movement had not previously possessed. The contrast with the Congress’s composite nationalism completes the field. The nationalism of Abul Kalam Azad and of the Congress mainstream located Indian nationhood in the shared anti-colonial struggle of all the land’s communities and refused every religious criterion of membership. Savarkar’s machine refuses the same criterion and installs another in its place: not belief but the geography of belief’s origins. The two answers agree that India is one nation; they divide on whether the nation’s unity survives the admission that its communities’ sacred maps point in different directions. Azad’s answer holds that the maps do not matter. The 1923 text holds that the maps are the matter. Indian politics has been the argument between those answers ever since, and the architectural reading insists that the argument be joined on the text’s actual ground rather than on the reputations the ground later acquired.

The Name Before the Definition: From Sindhu to Hindu

The book opens with a question about a word, and the opening is architectural rather than decorative. The first chapter, titled What Is in a Name, argues that names are not labels pasted onto peoples but condensed histories, and the chapters that follow excavate the name Hindu to its root in the river Sindhu. The excavation matters because the root fixes the definition’s foundations in geography before any criterion is stated. The peoples beyond the Indus, the Persians first in the author’s telling, rendered Sindhu as Hindu in their own phonetics, and the name travelled westward through Greek and Arabic usage as the designation of everyone who lived beyond the river, whatever those people believed. The Greeks dropped the initial sound and produced Indos and India. The land took its names from the river, and the people took their name from the land, in an order that Savarkar treats as the whole argument in embryo: first the territory, then the name, then the identity, and belief nowhere in the sequence.

The chapters on the name perform a second excavation, into antiquity. The author insists that the inhabitants of the land themselves adopted the exonym and made it a self-description, that the name Hindu appears in the inscriptions and literature of the land’s own rulers across centuries, and that the antiquity of the usage defeats the charge that the identity is a modern invention assembled for politics. The historical claims in these chapters are contestable in detail, and historians contest them: the epigraphic record for Hindu as a self-designation before the medieval centuries is thinner than the author’s rhetoric suggests, and the strongest scholarly caution, voiced across the literature Jaffrelot’s reader samples, holds that pre-modern usage of the term was primarily geographic and ethnic rather than national in the modern sense. The architectural reading registers the caution and sets it where it belongs. The definition does not depend on the antiquity argument succeeding, because the criteria operate without it; the excavation is foundation-laying rhetoric around a machine that would run on the criteria alone. Its presence in the book nevertheless reveals the author’s method, which is to present the classification as a discovery rather than an invention. Savarkar writes as a man recovering a fact that the land always contained, and the posture, half philology and half faith, is the posture Bakhle’s Savarkar and the Making of Hindutva (2024) identifies as the text’s literary engine: the definition presents itself as memory, and a definition received as memory binds more strongly than a definition received as proposal.

A chapter titled Hindus, a Nation carries the excavation into the definition’s first criterion. The argument runs that the people of the land constitute a nation in the objective sense the nineteenth century gave the word, bound by a common territory with natural frontiers, the Himalaya wall and the ocean moat, by a common descent the author traces through the commingling of peoples on the soil, and by a common civilisation whose festivals, laws and sacred places knit the territory into a single remembered home. The chapter on Sindhustan presses the territorial element to an intensity that later readers quote for its lyricism: the land from the Indus to the seas is addressed as a living mother and father in one, and the author’s prose rises to a pitch that sits oddly, to a secular ear, beside his atheism. The oddity dissolves under the architectural reading. The land in the text is not a goddess to be worshipped but a criterion to be satisfied; the lyric register performs attachment, the matribhumi element, in the act of describing it. A definition of belonging that could not be felt would fail as politics even if it succeeded as logic, and the chapters on the name exist to make the logic felt. Readers who charge the author with smuggling religion into a secular definition through the land’s sacral vocabulary identify a real tension in the prose, and the adjudication section returns to it; the tension is registered here as a tension of register, not of criteria, since the sacral vocabulary attaches to the land itself and never to a doctrine the candidate must affirm.

The Three Criteria: Pitribhumi, Matribhumi, Punyabhumi

The definition stands at the centre of the thirty-one chapters of Essentials of Hindutva, and the book’s second chapter announces its method in its heading: Hindutva is different from Hinduism. Before Savarkar defines anything he separates the word he is defining from the religion that shares its root, and he then builds the definition from three criteria that operate jointly. The first criterion is pitribhumi, the fatherland: the candidate’s ancestral line must trace to the land between the Indus and the seas, so that India is the land of the candidate’s forefathers and not a land of adoption alone. The second criterion is matribhumi, the motherland: India must be the candidate’s own territorial and political belonging, the country acknowledged as home in the first person, the land to which loyalty and obligation run. Commentators unpack the territorial criterion into these two elements, ancestry and belonging, because the text itself works with both: the culminating sentence of the book pairs the land regarded as fatherland with the land regarded as holy land, and the chapters on the name Sindhustan and on common blood supply the descent element that the fatherland term carries. The third criterion is punyabhumi, the holy land: the sacred geography of the candidate’s religion, its founders, its revelation and its pilgrimage, must lie within the same territory. A candidate who satisfies the land test and the descent test but whose religion locates its sacred events elsewhere fails the definition. All three criteria must hold together; the failure of any one excludes.

Savarkar’s own prose states the conclusion of the test in the book’s closing movement, where the author sums the argument in a single sentence that the Hindu Mahasabha subsequently adopted as its authoritative definition: a Hindu, the sentence runs, is a person who regards the land of Bharatvarsha, from the Indus to the seas, as fatherland and as holy land alike. The same chapters frame the three criteria in a second vocabulary, the vocabulary of nation, blood and civilisation. The text’s early chapters enumerate the essentials of Hindutva as a common nation or rashtra, a common race or jati whose blood the author traces to Vedic and post-Vedic ancestors on Indian soil, and a common civilisation or sanskriti expressed in shared festivals, laws, rites and historical memory. The two vocabularies describe one machine. Rashtra answers to the land criteria, jati to the descent criterion, sanskriti to the civilisational substance that the holy land criterion polices at the border of religion. The machine’s output is a classification of every population of the subcontinent, and the conceptual map below sets out the classification in the form the brief for this article requires: each criterion, the test it applies, the populations it passes and the populations it stops.

The map is the article’s findable artifact, and it is labelled here so that readers can cite it directly: Table 1, the three-criteria conceptual map of Savarkar’s 1923 definition, records the joint operation of pitribhumi, matribhumi and punyabhumi as the text constructs them.

Criterion	The test the criterion applies	Populations the criterion passes	Populations the criterion stops	Function inside the definition
Pitribhumi (fatherland)	Does the ancestral line trace to the land from the Indus to the seas?	Communities of Indian descent across every creed and sect	Populations whose ancestry the text treats as foreign in origin	Fixes membership in descent, independent of belief
Matribhumi (motherland)	Is India the candidate’s own acknowledged home and political belonging?	Residents who claim India as their country in the first person	Persons whose belonging the text assigns to another land	Converts descent into loyalty and obligation
Punyabhumi (holy land)	Does the sacred geography of the candidate’s religion lie within India?	Hindu, Buddhist, Jain and Sikh traditions born on Indian soil	Islam and Christianity, whose holy lands the text places in Arabia and Palestine	Polices the religious border without examining doctrine
The three jointly	Do all three criteria hold for the same person at the same time?	The Hindu of the definition, atheist or believer, of any caste or sect	Every candidate failing any single criterion	Produces the classification the whole book defends

Read the map from the bottom row upward and the book’s argumentative burden becomes visible. The joint row does the excluding, and it does the excluding almost entirely through the third criterion, because the first two criteria, as Savarkar applies them, pass nearly every long-settled population of the subcontinent. Indian Muslims and Indian Christians descend overwhelmingly from converts whose ancestors lived on the land for centuries, and most claim India as home without reservation. The definition therefore cannot exclude them on descent or on belonging without falsifying the demography its author elsewhere celebrates. The punyabhumi criterion carries the exclusion alone, and it carries it on a ground that has nothing to do with what Muslims or Christians believe: it locates the sacred events of their religions outside the territory and treats that location as disqualifying. Section by section, the text is candid about this. Savarkar writes that for Indian Muslims and Christians their holy land lies far away in Arabia or Palestine, that their mythology and godmen, their ideas and heroes, are not the children of Indian soil, and that consequently their love for the land faces a division his definition refuses to overlook. The candour matters for the adjudication later in this article, because the exclusion is architectural, stated in the open as a criterion, and not a matter of the author’s private prejudice leaking into a neutral design.

The map also displays what the definition conspicuously refuses to test. No row examines belief. No row asks whether the candidate accepts the Vedas, worships at a temple, follows a guru, observes a fast or professes any proposition at all. A row for doctrine would have been easy to write; nineteenth and early twentieth century religious classification in India was saturated with doctrinal tests, and the census operations of the colonial state, from the 1881 enumeration onward, wrestled publicly with the problem of defining a Hindu for counting purposes. Savarkar, who had read the census debates, declined the doctrinal route entirely and built a definition from land, blood and sacred geography instead. That refusal is the single most consequential architectural choice in the book, and its consequences run in two directions at once. Against the orthodox, it strips priests and scriptures of the power to excommunicate: no theological authority can expel a person who satisfies the three criteria, because theology is not among the criteria. Against Indian Muslims and Christians, it erects a border that no change of belief, loyalty or conduct can cross, because the border sits in the geography of a religion’s founding events, which no individual convert’s descendant can relocate. Readers who call the definition generous point to the first consequence; readers who call it exclusionary point to the second. Both are reading the same machine. The architectural reading insists only that the machine be described accurately before it is praised or condemned, and that its designer’s other commitments be tested against the machine’s logic rather than against the reputation the machine later acquired.

The borderline cases the text works through expose the machine’s finer gears, and three of them deserve record. The first concerns descent through mixed lines. The chapters on the commingling of peoples argue that the population of the land was formed by the fusion of stocks across millennia, that the blood of the candidate is therefore judged by its rootedness in the soil rather than by any fantasy of purity, and that children of the land’s fathers and mothers stand inside the definition by the pitribhumi criterion however their lines cross. The design intent is consolidationist: a descent test strict enough to exclude the mixed would have excluded nearly everyone, and the author, who had read his anthropology in the prison library, knew the demography too well to write it. The second case concerns the convert’s grandchildren. The text refuses to treat conversion as an event that ancestry can survive unchanged at the level of the definition, because the punyabhumi criterion attaches to the religion a person professes in the present, and the descendant of converts who professes Islam or Christianity fails the third test however many centuries the family’s tombs have stood in Indian earth. The harshness is deliberate and acknowledged: the author tells the reader that the gate is the gate, and that affection for the persons stopped at it cannot widen it. The third case concerns the Hindu who emigrates. A family of the land settled in Fiji, Trinidad or Natal carries pitribhumi in its ancestry and punyabhumi in its temples, and the matribhumi criterion, belonging acknowledged in the first person, does the delicate work: the text’s diaspora passages treat the overseas Hindu as a member whose belonging stretches across the sea to the motherland, a reading the author’s own London years among Indian students abroad prepared. Diaspora membership on those terms costs the definition nothing and gains it a worldwide constituency, which the author’s successors in the movement exploited with energy he did not live to direct.

One internal objection to the map must be faced inside this section, because it is the objection the adjudication inherits. The sanskriti criterion, common civilisation, can look like a belief test in disguise: a civilisation has contents, the contents include religious contents, and a candidate who repudiates the civilisation’s sacred narratives might seem to fail the third element as surely as a candidate whose holy land lies abroad. The text’s answer, visible in its treatment of the Charvakas and of the Buddha, is that sanskriti names a shared historical life rather than a shared creed: festivals kept or merely witnessed, epics heard in childhood, laws and customs absorbed from the soil, a calendar of memory rather than a catechism. The atheist who garlands no image still lives inside the civilisation’s time, speaks its languages and carries its stories, and the framework counts the carrying as enough. Whether the answer survives scrutiny is a question for the adjudication; the design intent is plain, and its plainness is what distinguishes the 1923 machine from the theological definitions it displaced.

The grammar of the culminating definition deserves a sentence-level reading, because the sentence’s verbs carry the architecture. The Hindu of the definition is a person who regards the land as fatherland and as holy land: the operative verb is an attitude toward territory, not an assent to propositions, and the sentence contains no verb of belief at all. Its span runs from the Indus to the seas, the Sindhu and the sagara of the text’s own geography, so that the definition’s map is drawn by a river and an ocean rather than by a creed’s jurisdiction. Translators and commentators have fought over the sentence’s rendering for a century, and the fights are instructive: render holy land as sacred soil and the sentence grows warmer, render it as the land of one’s religion and the sentence grows colder, yet the sorting output is identical under every rendering, because the criterion attaches to the location of the sacred and not to the intensity of the regard. A machine whose output survives its translation disputes has its meaning in its structure, and the structure is what the architectural reading reports. The political reading, which attends to the sentence’s warmth, reads a summons; the architectural reading, which attends to its grammar, reads a test. The sentence is both, in that order.

The civilisation criterion, sanskriti, carries a content that the text spells out through the geography of pilgrimage, and the spelling deserves record because it shows the punyabhumi test operating as lived practice rather than as abstraction. The sacred map the chapters celebrate is a network: Kashi on the Ganga and Rameshwaram at the land’s southern tip, Dwarka in the west and Jagannath Puri in the east, the twelve Jyotirlingas and the river confluences, a lattice of sites that no single sect owns and every region touches. A pilgrim walking the lattice walks the definition, binding the territory into one remembered home by foot and by story, and the author treats the lattice as the civilisation’s own cartography, drawn by devotion across centuries without a state’s instruction. The epics perform the same binding in narrative form: the Ramayana’s geography runs from Ayodhya to Lanka, the Mahabharata’s from Hastinapura to Kurukshetra and beyond, and the text claims both poems as the common property of every community the definition includes, the atheist’s inheritance as fully as the temple priest’s. This is the sense in which the framework’s cultural substance is territorial at root. The civilisation the third element names is not a list of propositions but a land made memorable, site by site and story by story, and the punyabhumi criterion simply asks on which land a given religion performed that making. Religions born on the soil did their remembering there; religions born elsewhere remember another soil, however faithfully their Indian adherents keep the festivals of their own tradition. The asymmetry is the definition’s crux, stated in the text’s own materials, and readers who would refute the framework gain nothing by refusing to see how deliberately the crux was constructed.

The Departure from Theology: Hindutva Is Different from Hinduism

The second chapter of the 1923 text performs the separation on which the whole framework rests, and it performs the separation with unusual explicitness. Hinduism, in Savarkar’s usage, names a religious system, one system among the several that arose on Indian soil, with its own theologies, rituals and schools. Hindutva names something larger and differently constituted: the text declares that the word covers the totality of the thought and activity of the Hindu people across their history, and it absorbs into that totality elements that no theology contains. A Hindu by religion might fail to be a Hindu by the definition if the land criteria failed, an outcome the text accepts without flinching; a person with no religion at all might satisfy the definition completely, an outcome the text not only accepts but insists upon, because the author’s own chapters on Buddhism praise the Buddha as a flower of the same civilisation and treat the Buddhist centuries as Hindu history in the civilisational sense.

The inclusion of the Indic religions follows mechanically. Buddhism, Jainism and Sikhism were founded on Indian soil by teachers whose sacred geography is Indian: the Buddha’s birth at Lumbini and enlightenment at Bodh Gaya, Mahavira’s career in the Gangetic plain, Guru Nanak’s birth at Talwandi and the Sikh sacred sites of Punjab. Their adherents therefore pass the punyabhumi test, and Savarkar says so in terms, counting Sikhs, Jains and Buddhists within the Hindu fold of the definition even where those communities maintain distinct religious identities and, in the Sikh case, a distinct political history. The definition thereby claims for its category millions of people who did not describe themselves as Hindus, a claim whose political uses later generations explored thoroughly. Architecturally, the point to register is the direction of the argument: the inclusion does not rest on shared doctrine, since Mahavira rejected the Vedas and the Buddha rejected the authority of sacrifice, and the text knows both facts. It rests on geography. The machine sorts by the map coordinates of sacred events, and the coordinates of Bodh Gaya and Talwandi fall inside the territory.

The exclusion of Islam and Christianity follows with the same mechanical quality, and the architectural reading requires that it be stated as plainly as the text states it. Savarkar’s argument is that conversion changed a family’s theology without changing its ancestry, and that the definition, honouring ancestry under the first two criteria, must nonetheless ask where the converted family’s sacred geography points. His answer points it to Mecca and to Jerusalem, and the definition stops there. The strongest version of the objection to this design, stated by Chaturvedi in Hindutva and Violence (2022), holds that the punyabhumi criterion functions in practice as a religious test wearing territorial dress, since it sorts populations exactly along the line between Indic and Abrahamic religions and assigns civic suspicion to one side of the line. The strongest version of the defence, stated across the Savarkar literature and given archival depth by Bakhle in Savarkar and the Making of Hindutva (2024), holds that the criterion measures attachment rather than creed: the question is not what a citizen believes but where the citizen’s civilisation locates its sacred centre, and a framework of national definition is entitled to ask the second question even where the first is forbidden. This article adjudicates between those versions in a later section; here the design fact is enough. The border of Savarkar’s Hindu runs between religions born inside the territory and religions born outside it, and belief is nowhere the border’s substance.

Two further populations expose the machine’s joints, and the text addresses both. The Parsis of Gujarat and Bombay arrived from Persia and locate their religion’s founding events in Iran; the Jews of Cochin and the Bene Israel of the Konkan locate theirs in Palestine. Savarkar treats both communities with respect in the text and in his later speeches, praises their loyalty to India, and still cannot admit them to the definition, because the punyabhumi criterion, applied honestly, stops them at the same gate that stops Muslims and Christians. He accepts the result. A definition that bends for sympathetic cases would not be a definition, and the author’s willingness to exclude communities he admired, on the same ground that excludes communities he opposed, is evidence that the criterion was an architectural commitment and not a discretionary weapon. The same honesty surfaces in the treatment of the atheist, the subject of the next section, where the machine’s indifference to belief reaches its most startling output: the author himself, a man who rejected God, scripture and idol, stands inside the identity by the definition’s own terms, while a devout Muslim farmer whose family had tilled the same village field for five hundred years stands outside it. Any reading of the 1923 text that cannot accommodate both halves of that sentence has not yet described the machine.

The separation gains force from the text’s inventory of what the word Hinduism is asked to cover. The chapters parade the sects and systems of the land, the Vedic ritualists and the Vedantic monists, the devotees of Vishnu and of Shiva, the Jains who deny the Vedas and the Charvakas who deny everything, the Sikhs with their scripture and the tribal communities with their groves, and the author draws the moral without mercy: no creed runs through the list, no prophet commands it, no single book binds it, and a definition of the people framed as a test of belief would have to begin by expelling most of the people it claimed to define. The word Hinduism itself, in the text’s handling, is a convenience of outsiders, a collective label pasted over a federation of systems that agree on little except the soil that bore them. That handling is why the framework’s author could treat religion with a freedom that scandalised believers: religion, in the architecture, is the civilisation’s wardrobe rather than its body, splendid, inherited and changeable, and the body persists through every change of dress. The sentence is the framework’s most quoted provocation in the secondary literature, and Bakhle’s reading in Savarkar and the Making of Hindutva (2024) treats the provocation as the author’s characteristic move, the reduction of the sacred to the civilisational, performed with a poet’s confidence that the civilisation can bear the weight the sacred carried.

The definition’s refusal of doctrine had a practical pedigree that the census history of British India supplies. From the first synchronous enumeration of 1881, the colonial census had struggled to state what a Hindu was, and the struggle is documented in the commissioners’ own reports: doctrine failed as a test because practice varied beyond every catechism, temple worship failed because millions of the counted never entered temples, and caste failed because the category plainly included communities the orthodox refused to feed. Census officials from J.A. Baines onward fell back on residual and territorial descriptions, counting as Hindu those whom the other named religions did not claim. Savarkar knew this administrative literature and inverted its embarrassment into a principle. Where the census treated the absence of a doctrinal test as a defect to be apologised for, the 1923 text treats it as the discovery itself: the category was never doctrinal, the administrators’ failure to find the doctrine was the evidence, and the land-descent-geography machine states positively what the census could only state by elimination. The point matters beyond antiquarianism, because the post-1947 republic inherited the census problem and answered it through the post-1947 plural-civilisational settlement context, where citizenship displaced religious classification as the state’s master category while the census went on counting religions it declined to define. Savarkar’s machine and the republic’s settlement agree on the negative, that belief cannot carry the definition of a people, and divide on the positive, where the machine enthrones sacred geography and the settlement enthrones the citizen.

The Indian Supreme Court handed the distinction a judicial afterlife that the architectural reading alone predicts. In the Hindutva judgments of 1995 and 1996, arising from election petitions against candidates whose speeches had invoked Hindutva, Justice J.S. Verma’s bench held that the word Hindutva, in its ordinary usage, denotes a way of life or a state of mind rather than the Hindu religion narrowly understood, and that Hinduism itself resists definition as a single creed with a single prophet and a single book. The Court’s language descends, by a route the judges did not need to trace, from the separation the second chapter of the 1923 text performs: Hindutva as the civilisational totality, Hinduism as one religious system inside it. Critics of the judgments, among them scholars of secularism whose objections the record preserves at strength, reply that the way-of-life reading launders a majoritarian political programme into a cultural description, and that a court which cannot define the religion should hesitate to define the civilisation. The dispute is live in the literature and is marked here as disputed. For the architecture, the narrower point suffices: a distinction drawn by a prisoner in 1923 proved robust enough to structure a Supreme Court judgment seven decades later, which is not the career of a slogan. It is the career of a definition that had been built, as its author built everything in those years, to survive confiscation and to travel without its maker.

The Text as History: The Civilisational Narrative of the Middle Chapters

A definition that sorts by sacred geography and shared memory must supply the memory, and the middle chapters of Essentials of Hindutva supply it in the form of a compressed civilisational history whose argumentative function repays attention. The narrative opens with the Vedic age on the banks of the Sindhu, follows the making of a people through the commingling of stocks the author insists upon against every purity doctrine, and arrives at the figure who organises its moral centre: the Buddha. The chapters on Buddhism are the surprise of the book for readers formed by later polemics. Savarkar treats the Buddha with reverence, counts the Buddhist centuries as a glorious internal chapter of the civilisation, and explains Buddhism’s decline on Indian soil through the political disarmament he believed the doctrine of non-violence induced in the face of invasion, reserving his severity for the consequence rather than the teacher. A chapter headed Reverence to Buddha stands in the table of contents of a book routinely described as a charter of Hindu supremacism, and the architectural reading accounts for the reverence without strain: the Buddha passes every criterion the machine applies, his sacred geography is Indian to the last site, and the civilisation that produced him is the substance the sanskriti criterion names.

The narrative then runs through the centuries of invasion and resistance that supply the definition’s emotional treasury: the Arab and Turkic incursions, the Rajput and Maratha resistances, and the culminating figure of Shivaji, whose seventeenth century state the author treats as the civilisation’s political self-assertion in arms. The function of the narrative inside the machine is precise. The punyabhumi criterion requires a sacred geography, and geography becomes sacred through remembered events; the middle chapters are the memory, arranged so that the map of the land and the map of the story coincide. Jaffrelot’s analysis in Hindu Nationalism: A Reader (2007) names the psychological engine of such narratives as the triad of stigma, fear and emulation: a community stigmatised as divided and effeminate answers by emulating the imagined unity of the other and narrating its own past as a single people’s epic. The architectural reading accepts the engine and contests the inference. That a narrative performs psychological work does not settle whether the classification it supports is well built; genealogies of motive explain why a definition was wanted, and the criteria decide what the definition, once wanted, actually says. The middle chapters also expose the narrative’s own exclusions, since the story’s heroes are chosen from the civilisation’s martial and Sanskritic registers and its centuries of Bhakti, Sufi exchange and regional vernacular flowering receive a fraction of the attention the wars receive. The selection is a fact about the machine’s fuel. It does not alter the machine’s gears, and the distinction between fuel and gears is the distinction this article exists to maintain.

The book’s most famous sentence stands near its threshold and states the narrative’s claim in nine words: Hindutva, the author writes, is not a word but a history. The sentence is the hinge between the definition and the middle chapters, and its logic completes the architecture. A word can be defined by stipulation, and a stipulation can be contested by a better stipulation; a history can only be narrated, and a narration binds its hearers by recognition rather than by argument. The middle chapters recruit the reader’s recognition, marching the remembered past across the land the criteria describe, until the reader who has followed the march feels the definition as the story’s summary rather than as the author’s proposal. Savarkar’s longer histories work the same seam at scale: Hindu Pad Padshahi of 1925, his Marathi account of the Maratha imperial century, narrates the civilisation’s political self-assertion as the recovery of a nationhood the invasions had interrupted, and the narrative habits of the history, its heroes and its arcs, are the narrative habits of the definition’s middle chapters enlarged. The adjudication between the two great readings must therefore handle the famous sentence with both hands. To the political reading, the sentence confesses the project: a history is being manufactured to order, and the definition is its instrument. To the architectural reading, the sentence states the sanskriti criterion’s content: the common civilisation just is the common history, carried in memory, and a definition that names the memory names the substance of the belonging. Both hands grip something real. The verdict of this article, that the architecture is prior, rests on the observation that the history is summoned to fill a criterion the definition had already fixed, as bricks are summoned to a wall whose plan precedes them.

The Atheist Inside the Framework: Savarkar’s Rationalism

The biographical fact that unsettles casual readers of the 1923 text is that its author did not believe in God. Savarkar described himself across his later writings as a rationalist and an atheist, he subjected Hindu religious practice to a utilitarian audit of a severity no missionary controversialist matched, and he held that doctrines stood or fell by their service to human welfare in this world. Readers formed by the political-ideological reading tend to file the atheism as an eccentricity, a private quirk irrelevant to the public doctrine. The architectural reading files it differently: the atheism is a prediction of the design. A definition of Hindu identity that tests land, descent and sacred geography, and that never tests belief, can be held with complete consistency by a man who believes nothing, and it can be held by no other kind of definition. Savarkar’s rationalism is the framework’s indifference to theology demonstrated on the framework’s own author.

The text itself prepares the demonstration. The chapters of Essentials of Hindutva treat the Charvakas, the materialist school of classical Indian philosophy, as members of the civilisational family, and they treat the heterodox systems with a respect that a theological definition could not have afforded, since several of those systems deny the Vedas, deny a creator and deny the soul. The author’s argument runs that a philosophy born on Indian soil and addressed to Indian questions belongs to the civilisation whatever it concludes, because belonging was never a matter of conclusions. Later, in the Ratnagiri internment, he converted the implication into a public position. The anti-superstition essays of the middle 1930s, preserved in the Ratnagiri-period volumes of the Savarkar Samagra Sahitya, attack the assignment of supernatural power to objects, places and animals, ridicule the commerce of priestcraft, and demand that every inherited practice present its credentials before the bar of utility. He wrote against the worship of the cow with a bluntness that cost him orthodox goodwill in his own district of internment, and he wrote against scriptural literalism with the same instrument: a text that sanctioned a harmful practice forfeited its authority over that practice, whatever its sanctity.

What did Savarkar say about cow worship?

Savarkar opposed cow worship outright and said so in essays that scandalised the orthodox of Ratnagiri. He defended the cow as a useful animal deserving protection on grounds of husbandry and national economy, and he treated the assignment of divinity to the animal as superstition that corrupted religion and policy together.

The rationalism extends to his handling of his own tradition’s foundational texts. Where a theological Hinduism treats the Vedas as revelation, Savarkar treats them as human documents of great antiquity and mixed value, to be honoured as ancestors are honoured and obeyed only where reason ratifies them. His celebrated sentence on the subject, repeated across the Ratnagiri writings, holds that the book that loses its relevance must be shelved with respect rather than followed into ruin. The defensible summary of the position, in the terms of the 1923 architecture, is that Savarkar transferred authority from scripture to the nation conceived as a civilisational body: the nation was the constant, its scriptures were instruments, and instruments were judged by output. That transfer is precisely what the definition of Hindutva enacts in the field of identity. The definition locates Hindu-ness in a body of people attached to a land and a civilisational memory, and it locates religious doctrine downstream of that attachment, as one expression among others of a civilisation that also produced atheists, logicians and kings.

The scholarly literature divides on the weight to give this material. Chaturvedi’s Hindutva and Violence (2022) reads Savarkar primarily as a theorist of political enmity and gives the rationalist essays a subordinate place in the argument, treating the strategic requirements of mobilisation as the key to the oeuvre. Bakhle’s Savarkar and the Making of Hindutva (2024), built on a close reading of the Marathi corpus that Anglophone commentary often skips, restores the essays and the poetry to the centre and presents a thinker whose authority rested on ascetic self-fashioning and literary labour as much as on doctrine. On the evidence of the 1923 text alone, the architectural reading sides with the restoration: the atheism and the rationalism are not decorations on the framework, they are load-bearing, because a theological definition in the hands of an atheist author would collapse under its own hypocrisy, and Savarkar, whatever else his critics establish, built a definition he could personally satisfy. The definition’s famous generosity toward heterodoxy is the author’s self-portrait generalised into a rule.

The rationalism had a forward edge as well as a critical one, and the edge shows in the author’s writing on science and machinery. Across the Samagra essays he celebrated the machine age, urged Indians to master the technologies of production and war without metaphysical trembling, and mocked the habit of consulting scripture on questions that engineering answers. His rule for such questions was uniform: a practice, a doctrine or a machine earns its place by its fruit in this world, and the appeal to antiquity counts for nothing against a demonstrated harm or a demonstrated use. The rule’s application to religion produced his most quoted provocations, among them the argument that the cow, protected by sentiment, should be protected instead by economics, and the corollary that a deity who requires protection from his devotees has confessed his own nature. The author’s instructions for his own end complete the picture the essays draw. He died on 26 February 1966 after renouncing food and medicine in the discipline he called atmarpan, self-offering, and the arrangements recorded by his associates for the disposal of his body dispensed with the apparatus of orthodox rite in favour of the crematorium’s efficiency. A theological definition of the man’s identity would have to posthumously expel him or explain him away. The 1923 definition absorbs him without a ripple, which is the closest thing to an experiment the history of ideas affords: the framework was tested on its own author, at the point of maximum strain, and held.

The constructive side of the rationalism appeared in his handling of festivals and public ritual, where the audit by utility produced reforms rather than abolitions. A festival, in the Ratnagiri writings, is a machine for assembling the community in public, and a machine is to be repaired, fuelled and aimed rather than smashed: the author advocated stripping festivals of their caste seating and priestly tolls while enlarging their civic content, so that the calendar of the civilisation did the consolidating work the definition required. The same logic governed his notorious interventions in language, where he coined and cleaned Marathi vocabulary with a rationaliser’s impatience for inherited clutter. Across domains the pattern holds: inherit nothing on authority, retain everything that works, and judge working by the standard of the consolidated nation’s strength. Readers who find the pattern cold have located its genuine cost. A civilisation audited by utility loses the protections that sanctity throws around the useless and the beloved, and the author’s own tradition felt the loss in his lifetime, in the orthodox boycotts his essays provoked. The architectural reading does not ask the reader to admire the audit. It asks the reader to notice that the audit is the definition’s twin: the man who sorted identity by land and memory rather than by belief was bound, by the same cast of mind, to sort practices by fruit rather than by pedigree, and the two sortings illuminate each other across the whole corpus.

Caste as the Framework’s Internal Enemy: The Ratnagiri Practice

The same architecture predicts the second biographical commitment that puzzles the political reading: Savarkar’s war on caste. Between his release from prison in January 1924 and the end of his internment in May 1937, confined to Ratnagiri district and barred from politics, he conducted a sustained campaign against the prohibitions that fragmented Hindu society in the Konkan. The campaign had a written charter. His essay of 1931 on the seven shackles of Hindu society enumerated the prohibitions to be broken: sparshabandi, the ban on touch across caste; rotibandi, the ban on eating across caste; betibandi, the ban on marriage across caste; vyavasayabandi, the assignment of occupation by birth; sindhubandi, the ban on crossing the sea; vedoktabandi, the reservation of Vedic rites to the twice-born; and shuddhibandi, the ban on readmitting converts to the fold. The list is the negative image of the 1923 definition. The definition imagines a single civilisational people sorted by land and memory; the shackles describe the machinery by which that people sorted itself into sealed compartments and lost the capacity to act as one body. Consolidation required the compartments to go.

The practice matched the charter, and the internment years supply dated instances in quantity. In 1925 Savarkar began work in the Maharwada, the Mahar quarter of Ratnagiri, organising joint devotional singing that seated ex-untouchable and caste Hindu worshippers in one assembly, and he campaigned through the towns of the district, Dapoli, Khed, Chiplun, Devrukh, Sangameshwar, Kharepatan, Devgad and Malvan among them, to end the seating of ex-untouchable children apart in schools, distributing slates and money to parents who sent their children to common classrooms. He organised inter-caste dining, sahabhojan, at event after event, treating the shared meal as a public referendum on rotibandi that the orthodox lost each time a dining hall filled. He defended shuddhi, the readmission of converts, as the demographic corollary of the definition: a framework that counted ancestry could hardly refuse return to the descendant of converts who sought it. B.R. Ambedkar, whose own war on caste proceeded from premises Savarkar did not share, wrote to him in terms the record preserves, conveying appreciation of the reform work and observing that the removal of untouchability would remain incomplete until chaturvarna itself was destroyed. The letter is evidence of the campaign’s reach, and its qualification is evidence of the distance between the two men’s premises, a distance the architectural reading keeps in view.

The shuddhi question, the seventh shackle’s concern, binds the caste campaign back to the definition with a knot worth untying in the open. Orthodox practice treated conversion out of the Hindu fold as irreversible: the door of exit stood open, the door of return was barred by the prohibition on readmission, and the demographic consequence across centuries of conversion was a one-way street the sangathan movement of the 1920s marched to reverse. Savarkar’s support for shuddhi, the ceremonial readmission of converts and their descendants, follows from the 1923 machine with the usual mechanical fidelity. The definition classifies by descent, land and sacred geography, and a descendant of converts who returns to a tradition of the soil alters the third element by his own act; a framework that refused the return would be policing a boundary its own criteria do not draw. The Ratnagiri practice therefore included public welcome ceremonies for the readmitted, conducted against orthodox protest, and the author wrote in defence of the practice with the argument his architecture supplied: a civilisation that cannot reabsorb its own children has confessed that its membership runs on rules other than the ones its definition states. The shuddhi paragraph of the campaign also carries the movement’s sharpest edge, since reconversion in the 1920s was a competitive practice aimed at populations the missionaries of two religions were courting, and the architectural reading does not soften the competition by naming its logic. It records the logic and the competition together, as the record requires.

What was the Patitpavan temple at Ratnagiri?

The Patitpavan temple at Ratnagiri, consecrated in 1931 with funds raised by the merchant Bhagoji Keer and with a Shankaracharya presiding, opened its sanctum to every Hindu community, including the ex-untouchable castes whom every other temple in the district barred. Savarkar promoted the temple as proof that exclusion was custom, not law.

The motivation question divides the literature, and the architectural reading adjudicates it by the framework’s logic. Savarkar’s anti-caste campaign was not Gandhian, since it owed nothing to a doctrine of the oppressor’s moral conversion, and it was not Ambedkarite, since it sought the abolition of caste compartments in the interest of the consolidated nation rather than the annihilation of the religious order that sanctified them. It was instrumental in the precise sense the framework dictates. Caste exclusion expelled from common life millions of people whom the 1923 definition counted as Hindus by land, descent and sacred geography; the definition’s own author could watch his category contradicted daily by the dining rules and temple doors of his district. The campaign to break the shackles was the framework defending its classification against the society the classification described. Critics who reply that instrumental reform evaporates when the instrument is no longer needed state a genuine risk, and the risk attaches to the record of the movement after Savarkar; the architectural point stands that within the 1923 design, anti-caste practice is not an ornament but a theorem.

Why did Savarkar attack the ban on crossing the sea?

Savarkar attacked the sea-voyage prohibition as arithmetic folly: a ban that kept Hindu merchants and students from crossing the ocean surrendered trade, learning and naval skill to others. The prohibition, sindhubandi, stood on his 1931 list of the seven shackles that a consolidated Hindu society had to break for reasons of survival.

The campaign’s texture deserves a paragraph of its own, because the Ratnagiri record answers the charge that the anti-caste work was a pamphleteer’s posture. Ratnagiri was a bastion of orthodoxy, dominated by Chitpavan and Deshastha Brahmin establishments whose authority the internecine challenged daily and in person. The methods were deliberately theatrical and deliberately repeated: the joint bhajan sessions that opened the Maharwada to caste Hindu singers and the town’s temples of song to Mahar voices; the public meals at which the author ate with ex-untouchable guests before cameras of the local press; the school inspections that exposed headmasters who reported integrated classrooms and practised segregated ones; the reform of the public Ganapati festival in the district, which Savarkar’s circle reorganised so that the procession and the worship admitted every caste to the platform. Each method attacked a named shackle from the 1931 list, and each was chosen for visibility, since a prohibition enforced by social boycott dies only when the boycott is publicly defied and seen to fail. The orthodox response, social and economic pressure on participants, is documented in the biographies, Dhananjay Keer’s authorised life of 1950 among them, and the campaign’s continuation across thirteen years of internment, long past any plausible horizon of political reward, is the record’s answer to the tactical reading of the reform work. Tactics expire when the audience leaves; the Ratnagiri audience was a district town, the reward was nil, and the campaign ran from 1924 to the internment’s end in 1937 without a pause the record discloses.

The Ambedkar comparison fixes the campaign’s coordinates in the decade’s reform geography, and the comparison must be drawn with care because the two men agreed on a premise and divided on everything built upon it. In Annihilation of Caste, the address prepared in 1936 for the Jat-Pat Todak Mandal of Lahore and published by its author when the invitation was withdrawn, B.R. Ambedkar argued that caste is not merely a division of labour but a division of labourers, sanctified by scriptural authority, and that no reform which leaves the sanctity intact can reach the root. Savarkar’s seven shackles essay of 1931 agrees that scriptural sanction is the enemy and refuses the sanctity with equal bluntness; the two men’s published positions on the authority of the shastras are closer than either camp’s later mythology admits. The division opens at the destination. Ambedkar’s argument runs toward the destruction of the religious order that houses caste, and his Yeola declaration of October 1935, that he had been born a Hindu but would not die one, announced the exit his logic required and his conversion of 1956 completed. Savarkar’s argument runs toward the consolidation of the order’s population under a definition that makes caste compartments unintelligible, and it requires the order to survive, emptied of its internal fences. One man sought to annihilate caste by leaving the house; the other, to annihilate caste by demolishing the house’s partitions and keeping the house. The 1923 definition underwrites the second project exactly, since a category built on land and memory has no internal walls to defend, and it cannot underwrite the first, since exit from the religion leaves the definition’s criteria untouched and the exiting population still, by the machine’s cold arithmetic, classified. The sentence will offend, and it is the machine’s sentence rather than this article’s: Ambedkar himself, by pitribhumi and by the Indian origins of the faith he finally embraced at Nagpur in 1956, never left the definition for a day. The offence the sentence gives is the measure of the distance between a politics of dignity and a politics of consolidation, and the architectural reading reports the distance without pretending to close it.

The Mahasabha Years and the Organisational Vacuum

The framework’s incompleteness is organisational, and the incompleteness is visible in the biography the moment the internment ends. Released unconditionally in May 1937, Savarkar assumed the presidency of the All-India Hindu Mahasabha at its Ahmedabad session in December 1937 and held the office through re-election until 1943. The Mahasabha was the natural vehicle to hand: the organisation had adopted his definition as its authoritative statement of who a Hindu was, its leadership included men like Bhai Parmanand, the Lahore leader whose conception of a consolidated Hindu nation predated the book and whose correspondence with the interned author, preserved among the Savarkar Papers at the National Archives of India, documents the exchange of ideas between the two men across the Ratnagiri years. Yet the vehicle and the definition fitted each other imperfectly. The Mahasabha was a body of notables, strong in resolutions and weak in cadres, and a definition is not a membership drive. The organisation could declare who counted as a Hindu; it could not drill, feed and discipline the counted into an instrument.

The presidential addresses of 1937 to 1943, collected as Hindu Rashtra Darshan, show the author working the framework in public and meeting its costs. At Ahmedabad in December 1937 he told the session to face facts and stated that two antagonistic nations lived side by side in India, a sentence the political-ideological reading quotes as the definition’s confession and the architectural reading quotes as the definition’s arithmetic: a classification that sorts populations by sacred geography will, at the level of politics, describe two nations wherever the sorting line coincides with a communal boundary. Through the war years he pressed the slogan of Hinduising politics and militarising the Hindu body, urged Hindu recruitment into the armed forces against the Congress policy of non-cooperation with the war effort, and conducted the politics of responsive cooperation with provincial ministries. The addresses are political documents, and the strongest political reading of Savarkar rests its case on them. The architectural reading absorbs the case rather than denying it: the addresses operate inside the 1923 classification, drawing boundaries the classification drew, and they display the classification’s strengths and cruelties in application. What they never do is amend the definition, because the definition was load-bearing for everything built on it.

How many terms did Savarkar serve as Hindu Mahasabha president?

Savarkar presided over the Hindu Mahasabha for seven consecutive years, elected at the Ahmedabad session of December 1937 and re-elected through the sessions that followed until 1943. The tenure covered the organisation’s most ambitious phase, its negotiations over the war, the Cripps offer and the shape of a post-war settlement.

The decline that followed had structural causes the biography alone does not explain. After 1943, with Savarkar’s health failing and his presidency ended, the Mahasabha lost its claim on the constituency the definition had mapped, and the assassin’s bullets of 30 January 1948 completed the ruin. Nathuram Godse had moved in Mahasabha and Sangh circles; Savarkar was arrested, tried at the Red Fort as a co-accused in the conspiracy to murder Mahatma Gandhi, and acquitted on 10 February 1949 by Justice Atma Charan, who found the approver Digambar Badge’s testimony against him wanting in corroboration. The Kapur Commission, appointed in 1966 and reporting in 1969, three years after Savarkar’s death on 26 February 1966, recorded conclusions adverse to him on the basis of testimony the trial court had not credited, and defenders of Savarkar contest the commission’s procedure and its findings to this day; the dispute is marked here as disputed, and neither the acquittal nor the commission settles it for honest readers. What the episode settles is the organisational question. The Mahasabha never recovered, and the constituency the definition described found its durable vehicle elsewhere, in the cadre organisation that K.B. Hedgewar had founded at Nagpur in 1925 on principles of daily discipline that owed the definition much and the Mahasabha little. That organisation’s story belongs to a later article in this series and enters the argument here solely as the measure of the vacuum: a framework of identity, however precisely engineered, does not march.

The strongest exhibit in the political reading’s file belongs in this section, because it arises from the Mahasabha years and tests the framework where the framework met power. In December 1941 the Hindu Mahasabha entered the coalition ministry of A.K. Fazlul Haq in Bengal, with Syama Prasad Mukherjee as finance minister, in a government sustained alongside the Muslim League’s provincial organisation; Mahasabha members likewise joined coalition arrangements in Sindh and in the North-West Frontier Province in the same war years, in provinces where the League’s presence in the legislature made such arithmetic unavoidable for any ministry at all. The spectacle of Savarkar’s party governing with the League, the organisation of the two-nation theory’s Muslim half, is quoted by critics as the framework’s hypocrisy exposed: the definition that declared two antagonistic nations in 1937 cohabited with the other nation in office within four years. The architectural reading answers on the record. A classification of identity is not a rule of alliance, and the 1923 text contains no clause governing coalition arithmetic; Savarkar defended the Bengal arrangement in the vocabulary of responsive cooperation, as the price of keeping ministries out of worse hands and of advancing Hindu recruitment and organisation under wartime conditions. Readers may judge the defence unconvincing as politics. As architecture it is consistent, and the consistency cuts both ways: a framework that permits its holder to ally with the classified other for provincial advantage has plainly ceased, at that moment, to function as a programme of enmity, whatever its function as a map of identity remained. The complication the brief for this article requires to be faced, that the later political activities carry substance the architectural reading must not wave away, stands faced: the substance is real, the addresses and the coalitions are political acts with political consequences, and the architectural claim survives them because it was never a claim that the author abstained from politics. It is a claim about what his definition, built in 1923 before the politics resumed, actually contains.

Two Readings of One Text: The Adjudication

The named disagreement can be adjudicated, on the assembled evidence, as follows. The political-ideological reading, in the strong form Chaturvedi gives it in Hindutva and Violence (2022), holds that the 1923 text is best understood as the founding charter of a political project: the definition exists to constitute a Hindu nation against an internal other, the punyabhumi criterion exists to mark that other, and the framework’s consequences in mobilisation, from the Mahasabha presidency to the Ayodhya decades, reveal the design’s intention. The architectural reading, defended in this article and supported at the level of the text’s construction by Bakhle’s Savarkar and the Making of Hindutva (2024), holds that the text is best understood as a conceptual machine for classifying identity: its criteria are land, descent and sacred geography; its refusals, of doctrine as a test and of theology as an authority, are deliberate design decisions; and the political consequences, however grave, are applications of the machine by actors who did not redesign it. Jaffrelot’s Hindu Nationalism: A Reader (2007) mediates between the two in the classroom, presenting the text inside a genealogy of nationalism while reproducing the definitional chapters at a length that lets students see the machine for themselves.

Three tests decide between the readings, and the architectural reading passes each test on evidence the political reading must set aside. The first test is predictive reach. A reading of a framework earns its standing by explaining the framework-holder’s other commitments without special pleading. The architectural reading predicts the atheism, since a definition innocent of doctrine can be authored by an atheist in good conscience; it predicts the anti-superstition essays, since a civilisation defined by land and memory may audit its own rituals by utility; it predicts the seven shackles campaign, since compartments inside the category contradict the category; it predicts the inclusion of Buddhists, Jains and Sikhs, the exclusion of sympathetic Parsis and Jews, and the tortured consistency of the exclusion of Indian Muslims and Christians on a single stated ground. The political reading predicts the Ahmedabad sentence of December 1937 and the Mahasabha addresses, and then must treat the atheism as camouflage, the temple-entry campaign as tactics and the Parsi exclusion as an embarrassment, none of which the record supports without strain.

The second test is fidelity to the text’s own order of operations. The book defines before it prescribes; its peroration is a definition, not a manifesto; its author’s prison conditions explain the priority, since definition was the only political act his jailers left open. A reading that treats the definition as a disguise for the politics must explain why the disguise occupies the whole book and the politics almost none of it. The third test is counterfactual discipline. Ask what the text would look like if the political reading were correct in its strong form, and the expected artefact is a programme: organisation, enemies, stages, demands. Ask what the text would look like if the architectural reading were correct, and the expected artefact is a classification with borderline cases worked through in the open, which is the artefact on the shelf, eighty-eight pages of it, reprinted under a question for a title. The adjudication follows. The political-ideological reading describes the framework’s career accurately and its design inaccurately; the architectural reading describes the design accurately and thereby explains the career, including the parts of the career, the Ahmedabad sentence and the Mahasabha’s two-nation arithmetic, that the career’s chroniclers quote most. Analytically prior means exactly this: the architecture first, because the politics ran on rails the architecture laid.

A fourth test, of method rather than of evidence, confirms the verdict and disciplines its scope. Legal theory distinguishes a rule from the intentions of its drafter: a test in a statute sorts the cases that arise under it, including cases the drafter never imagined, and the sorting is the rule’s meaning in operation. The 1923 definition behaves as such a test behaves. It sorted its author’s own unbelief, a case no drafter of a religious definition could have intended; it sorted the Parsis and the Jews against its author’s sympathies; it sorted Ambedkar, across his exit and his conversion, without consulting his wishes; and it sorted the diaspora, the converts’ grandchildren and the mixed lines by criteria stated in advance of the cases. A political programme does none of this work, because a programme addresses its friends. The definition addresses everyone, friend and stranger and enemy, in the same clauses, and the impersonality is the surest signature of its kind. The adjudication therefore closes with a clarification of what has and has not been decided. Decided: that the architectural reading describes the 1923 artefact accurately, predicts its author’s commitments, and explains the political career as application. Not decided: any question of the framework’s justice, wisdom or fitness for the republic that eventually governed the land the framework mapped. Those questions belong to the normative argument that Indian politics conducts without pause, and a reading that clarifies the artefact does that argument the service of giving it a fixed object to be about.

The political reading’s strongest form deserves its strongest statement before the section closes, because an adjudication that defeats a weak opponent proves nothing. In that form, pressed across Hindutva and Violence (2022), the argument holds that the definition was never an innocent taxonomy: written in the aftermath of the Moplah rising and amid the Khilafat mobilisation, addressed to a Hindu public its author wished to consolidate against a rival consolidation, the text intervenes in the politics of its hour with every clause, and the punyabhumi criterion in particular was drafted with its target in view, since no other clause was needed to produce the exclusion the hour demanded. The architectural reading concedes the intervention and denies the inference. A definition may be drafted for an hour and still be a definition, as a census schedule drafted for an administration still classifies; the question the adjudication must answer is not why the machine was built but what the machine, once built, does, and the machine does more, and other, than the hour required. The hour did not require the inclusion of atheists, the protection of heterodoxy, the exclusion of the admired Parsis, the abolition of the dining bans or the subordination of scripture to utility, and a reading that explains the text by its hour leaves those outputs as residue. A reading that explains the text by its architecture explains the hour’s outputs and the residue in a single account. That is the whole of the verdict, and it is enough: priority belongs to the reading that leaves less of the evidence unexplained.

The disentanglement can be tested once more against the first and most consequential act of inheritance. M.S. Golwalkar’s We or Our Nationhood Defined of 1939 opens in acknowledged proximity to the 1923 definition, adopts its territorial frame and then rebuilds the interior: where Savarkar’s machine sorts by criteria any classifier can apply, Golwalkar’s framework apprehends the nation as an organic body whose unity is felt in blood and culture and proved by the assimilation or subordination of those who stand outside it. The differences are architectural and they compound downstream: the atheist whom Savarkar’s criteria shelter becomes, in the organic framework, a problem of allegiance rather than a citizen of the definition; the caste compartments Savarkar attacked become, in the organic idiom, organs to be harmonised rather than shackles to be broken; and the classification’s impersonality gives way to a doctrine of the national organism in which the classifier’s judgement about essences does the work the criteria once did. The pillar’s later articles take up the second and third variants in their own right. For the foundational text the moral of the comparison is the one the architectural reading predicts: read the 1923 machine through its 1939 rebuild and every joint is mislocated, because the rebuild replaced the machine’s logic while keeping its vocabulary. Disentanglement begins by refusing the substitution.

What the 1923 Framework Committed To, and What It Did Not

A framework earns disentanglement from its descendants only when its commitments are stated with precision, and the 1923 text permits precision. The framework committed its holder to a Hindu identity constituted by land, descent and sacred geography, in which doctrine had no vote. It committed its holder to the inclusion of every Indic religious tradition and of every unbeliever born inside the classification, and to the exclusion of every adherent of the religions founded beyond the territory, however deep that adherent’s ancestry or however fervent that adherent’s patriotism. It committed its holder to war on the caste compartments that contradicted the classification from within, and to a rationalist audit of inherited practice by the standard of utility. It committed its holder to nothing else in particular: not to a party, since the Mahasabha adoption came from the organisation and not from the text; not to a state form, since the book’s political vocabulary runs to nation and civilisation rather than to constitutions; and not to a theology, since the text expressly dissolves the category of belief into the larger category of a people’s historical life. The later variants that this series separates, the organic-social framework Golwalkar built inside the Rashtriya Swayamsevak Sangh and the electoral idiom of the Bharatiya Janata Party decades afterward, inherited the classification and added the machinery the classification lacked. Judging the additions by the inventor’s standard, or the invention by the additions’ record, confuses the genealogy the series exists to keep straight.

The language question shows the classification working after its author’s death and inside other men’s arguments. Savarkar wrote first in Marathi, published his definition in English, and advocated Hindi in the Devanagari script as the link language of the nation he imagined, three positions that later politics fused into a single slogan. The fusion, and the resistance the fusion met in the Dravidian south and the federal republic, returns as the Hindi-Hindutva-Hindostan triad, where the triad’s coherence is the question under adjudication. Against that later fusion the 1923 text maintains its distinctions: language enters the framework as an element of sanskriti, of shared civilisation, and an element is not a criterion. A Tamil-speaking Hindu of Madurai satisfies the definition without a word of Hindi, and a Hindi-speaking Muslim of Lucknow fails it on the punyabhumi clause alone. The precision offends both camps of the language wars, which is the usual signature of an architectural fact surviving its political uses.

The constitutional afterlife of the definition runs on the same precision. The Constituent Assembly of 1946 to 1949 confronted the religion-state question the definition had answered in its own way a generation earlier, and the Assembly answered it otherwise, building the settlement whose tensions are traced in the religion-state founding tension. The comparison is instructive because the two answers share a refusal: the Assembly, like Savarkar, declined to define Indian belonging by doctrinal belief, and built citizenship on territory and law instead. Where the answers divide is the punyabhumi clause, which no constitutional order founded on equal citizenship could admit, since the clause sorts citizens by the geography of their religions’ founding events. The division explains the definition’s permanent position in Indian politics: available as a theory of the nation, unusable as a rule of the state, and endlessly cited by movements negotiating the distance between those two stations. That negotiation, recast for a later century, is the subject of the integrated three-tensions framework, and the comparator for the founding generation’s own handling of charismatic definition-makers stands in the founding triangulation context, where three men who rejected Savarkar’s answer built the institutions that outlasted it.

The sources for the argument of this article are named here in the order of their authority. The primary text is Essentials of Hindutva (1923), read in the full and in the definitional chapters with the attention to chapter order that the adjudication required. Behind it stand the Ratnagiri-period writings preserved in the Savarkar Samagra Sahitya, the ten-volume collected works issued by Hindi Sahitya Sadan, where the anti-superstition essays and the seven shackles charter supply the framework’s own commentary on itself; the Hindu Mahasabha resolutions and presidential addresses of 1937 to 1944, preserved at the Nehru Memorial Museum and Library, which record the definition’s application by its author in office; and the Ratnagiri correspondence with Bhai Parmanand and others, preserved among the Savarkar Papers at the National Archives of India, which documents the intellectual engagements that produced the conceptual choices. Readers who consult the collected works in digital form, or who receive the archival transcripts as office documents from a reading room, can open and read the files in a browser through an office file viewer for Excel, DOCX and PPTX documents, a small practical point that matters because the Marathi corpus, on which the architectural reading leans, circulates chiefly in such files. The secondary adjudication rests on Bakhle (2024) and Chaturvedi (2022), with Jaffrelot’s reader (2007) as the classroom standard against which both were checked.

The claim of this article can therefore be restated in its final form, stripped of ornament. Savarkar’s 1923 framework was territorial, ancestral and civilisational, not theological; the conceptual-architectural reading is what makes the framework’s specific commitments intelligible; and the disentanglement of the 1923 text from the Hindutva variants that followed it begins, as this article has begun, with the machine on the page rather than with the crowds in the street. The crowds matter, and the series will reach them in their order. The machine came first, in a cell at Ratnagiri, built by a prisoner with smuggled paper, and the machine’s design, generous and cruel in the same joints, is the fact of 1923 that every later fact presupposes.

One commitment of the framework remains to be stated in its own right, because the later variants inherited it and because the disentanglement fails without it. The framework is a theory of the nation that assigns the state no theological office and the citizen no religious test, and it thereby stands at a measured distance from the theocratic readings its name attracts in casual usage. Within the 1923 text the state’s business is the nation’s strength, the nation’s substance is civilisational, and the civilisation’s religions are its protected inheritance rather than its constitution. The distance should not be overstated into a liberalism the author did not hold: the framework’s nation has a civilisational character, majorities within it carry the memory that defines it, and minorities classified outside the definition face the civic coldness the classification implies, a coldness the author’s Ahmedabad sentence of 1937 stated without softening. The precision the architectural reading owes the record is the precision of the joints: where the framework separates, at the line between nation and state and between civilisation and creed, and where it fuses, at the line between the citizen and the sacred geography of the citizen’s faith. Later Hindutva variants moved both lines, Golwalkar’s organic framework fusing nation and social body more completely, the electoral variant bargaining the lines constituency by constituency. The 1923 text holds its lines where its author drew them, in a jail, with the century still young, and the lines are the legacy every successor had to accept, amend or evade.

Frequently Asked Questions

Q: What is Savarkar’s Essentials of Hindutva?

Essentials of Hindutva is the eighty-eight page English pamphlet that V.D. Savarkar composed in Ratnagiri jail and published at Nagpur in 1923 under the pseudonym A Maratha. Across thirty-one chapters the book defines who counts as a Hindu, separates the concept of Hindutva from the religion of Hinduism, and grounds the definition in land, ancestry and sacred geography rather than in belief. The 1928 reprint carried the title Hindutva: Who Is a Hindu?, and the Hindu Mahasabha adopted the book’s culminating definition as its authoritative statement of Hindu identity.

Q: What are pitribhumi, matribhumi, and punyabhumi?

The three terms name the three criteria of Savarkar’s 1923 definition, and all three must hold together. Pitribhumi, the fatherland, tests descent: the candidate’s ancestral line must trace to the land from the Indus to the seas. Matribhumi, the motherland, tests belonging: India must be the candidate’s own acknowledged home and the object of political loyalty. Punyabhumi, the holy land, tests sacred geography: the founding events of the candidate’s religion must lie within India. A candidate who passes the first two tests and fails the third falls outside the definition, which is the mechanism by which the text excludes Indian Muslims and Christians. The book’s culminating sentence pairs the first and third criteria, fatherland and holy land, in a single line that the Hindu Mahasabha later adopted verbatim as its official definition of a Hindu.

Q: Was Savarkar an atheist?

Yes, by his own repeated description. Savarkar called himself a rationalist and an atheist, rejected belief in a creator God, treated the Vedas as human documents rather than revelation, and subjected inherited religious practice to a utilitarian audit in the Ratnagiri-period essays preserved in the Savarkar Samagra Sahitya. The position was not a private eccentricity. His 1923 definition tests land, descent and sacred geography and never tests belief, so the framework’s author could satisfy his own definition in complete consistency, and the text’s hospitality to Charvaka materialists and other heterodox schools generalises the author’s case into a rule of the framework.

Q: What did Savarkar say about caste?

Savarkar held caste by birth to be a curse on the Hindu body and organised the Ratnagiri internment years, from 1924 to 1937, as a campaign against it. His 1931 essay on the seven shackles listed the prohibitions to be broken: bans on touch, on inter-caste dining, on inter-caste marriage, on occupations outside one’s birth caste, on sea voyages, on Vedic rites for the lower orders, and on readmitting converts. He organised inter-caste dining, promoted the Patitpavan temple of 1931, which admitted ex-untouchable worshippers, and defended the eradication of untouchability as a condition of national consolidation rather than as charity.

Q: How is Savarkar’s Hindutva different from Hinduism?

Hinduism, in the usage of the 1923 text, names a religious system: a body of theologies, rituals and schools that arose on Indian soil. Hindutva names the larger civilisational totality of the Hindu people across their history, in which religion is one department among others that include land, ancestry, law, festival, language and historical memory. The second chapter of the book performs the separation in its heading. The practical consequence is that a person may be a Hindu of the definition with no religion at all, as the atheist author was, and a devout practitioner of a religion founded outside India may fall outside the definition entirely.

Q: When did Savarkar write Essentials of Hindutva?

Savarkar formed the leading arguments during his decade in the Cellular Jail at Port Blair from 1911 to 1921, where the front matter of the 1942 edition records that released prisoners carried the definition back to India in memory between 1917 and 1919. He wrote the text out in the district jail at Ratnagiri in 1922, after his repatriation to the mainland in May 1921, and the manuscript was smuggled from his cell in May 1923 to Vishwanathrao Kelkar, who published the pamphlet at Nagpur in 1923. The reprint of 1928 gave the book its enduring title, Hindutva: Who Is a Hindu?

Q: Why did Savarkar exclude Muslims and Christians from Hindu identity?

The exclusion operates through the punyabhumi criterion, the third test of the definition. Savarkar conceded that Indian Muslims and Christians descended from ancestors rooted in the land and that India was their home, which passes the fatherland and motherland tests. He held that their holy land nevertheless lay outside India, in Arabia for Islam and in Palestine for Christianity, and that a sacred geography centred elsewhere divided the loyalty his definition required to be whole. Critics such as Vinayak Chaturvedi in Hindutva and Violence (2022) reply that the criterion functions as a religious test in territorial dress. Defenders answer that it measures attachment rather than creed. The text itself states the mechanism without disguise, and the same mechanism explains the definition’s treatment of the Parsis and the Jews, communities the author praised in print and nevertheless excluded on the identical ground of a holy land beyond the territory.

Q: Why did the 1923 edition of the book carry the name A Maratha?

Savarkar was a serving prisoner in Ratnagiri jail in 1923, transported for two life terms and forbidden the ordinary rights of an author. A title page bearing his name would have exposed the pamphlet to suppression and its publisher, Vishwanathrao Kelkar of Nagpur, to prosecution, and it would have cost the author what remained of his conditional liberties. The name A Maratha solved the problem: it identified the writer by region and by the martial tradition of Maharashtra without naming him, it carried a resonance the book’s first readers recognised, and editions printed after his release restored the authorship that the first edition had to hide.

Q: Did Savarkar count Sikhs, Jains and Buddhists as Hindus?

Yes. The 1923 definition counts the adherents of every religion founded on Indian soil as Hindus of the definition, because their sacred geography passes the punyabhumi test: the Buddha’s birth at Lumbini and enlightenment at Bodh Gaya, Mahavira’s career in the Gangetic plain, Guru Nanak’s birth at Talwandi and the Sikh sacred sites of Punjab all lie within the territory. The text praises the Buddha as a flower of the civilisation and treats the Buddhist centuries as part of the Hindu civilisational record. The inclusion rests on geography, not doctrine, since Buddhism and Jainism reject Vedic authority, and it describes the definition’s classification rather than those communities’ accounts of themselves.

Q: What part did the Andaman Cellular Jail play in forming Savarkar’s ideas?

The decade from 1911 to 1921 in the Cellular Jail at Port Blair was the definition’s long gestation. Denied writing materials, Savarkar composed in memory, taught fellow prisoners, and worked out the arguments that Ratnagiri later received on paper; the 1942 edition’s front matter records that prisoners released between 1917 and 1919 carried the gist of the treatise and the definition itself to India by oral transmission. The jail also taught a structural lesson the book embodies: institutions, presses and parties can be confiscated, while a definition, memorised and portable, survives transportation. The architecture of the 1923 text is the work of a man designing for confiscation.

Q: What was Savarkar’s position on the Manusmriti?

Savarkar treated the Manusmriti as a human document of its age rather than as revelation, and he attacked its caste injunctions directly. In the seven shackles essay of 1931 and in the Ratnagiri speeches he argued that scriptures function as instruments in the hands of those whom the arrangements of the day advantage, and that a text sanctioning the humiliation of one Hindu by another forfeited its authority over conduct. His general rule, repeated across the Samagra writings, held that inherited books earn obedience only where reason ratifies them, and that a book which has lost its relevance should be shelved with respect rather than followed into ruin.

Q: Who was Bhai Parmanand, and what was his connection to Savarkar?

Bhai Parmanand was the Lahore-based Hindu Mahasabha leader and Arya Samaj figure who had argued for the consolidation of Hindus into a single national body before Savarkar wrote a line of the 1923 text. Transported to the Andaman Islands in the Lahore conspiracy proceedings of 1915 and released in 1920, he shared with Savarkar the experience of the Cellular Jail and the vocabulary of Hindu organisation. During the Ratnagiri internment the two men corresponded, and the letters, preserved among the Savarkar Papers at the National Archives of India, document the exchange through which the older leader’s ideas met the younger prisoner’s definitional project.

Q: What happened when Savarkar was tried in the Gandhi assassination case?

After the assassination of Mahatma Gandhi on 30 January 1948, the Government of India prosecuted Savarkar at the Red Fort trial as a co-accused in the conspiracy, on the strength of his association with Nathuram Godse and the approver Digambar Badge’s testimony that Savarkar had blessed the departing conspirators. On 10 February 1949 Justice Atma Charan acquitted him, holding Badge’s evidence against him uncorroborated on the decisive points. The Kapur Commission, reporting in 1969 after Savarkar’s death, recorded adverse conclusions on testimony the trial court had not credited, and Savarkar’s defenders contest the commission’s procedure and findings; the question stands disputed in the literature.

Q: How do Janaki Bakhle and Vinayak Chaturvedi read Savarkar differently?

Chaturvedi, in Hindutva and Violence: V.D. Savarkar and the Politics of History (2022), reads Savarkar as a theorist of political enmity whose history-writing organised the Hindu nation against an internal other, and he weighs the 1923 text by the violence of the tradition that claimed it. Bakhle, in Savarkar and the Making of Hindutva (2024), reads the full Marathi and English corpus with philological care and presents the making of the concept as an act of literary and ascetic labour, in which the definition’s construction is the event to be explained. The first weighs the text by its career; the second reconstructs its architecture. This article’s adjudication follows the architecture while recording the career in full.

Q: How did Savarkar’s definition treat Jews and Parsis living in India?

The definition stops both communities at the punyabhumi gate, and Savarkar accepted the result in the open. The Parsis of Gujarat and Bombay, resident in India for a millennium, located their religion’s founding events in Iran; the Jews of Cochin and the Bene Israel of the Konkan located theirs in Palestine. The text and the later speeches praise both communities for loyalty to India and for their contributions to its life, and the definition nevertheless excludes them, because a criterion that bends for sympathetic cases ceases to be a criterion. The episode is the strongest internal evidence that the exclusion of Muslims and Christians was architectural rather than discretionary.

Q: What is the Savarkar Samagra Sahitya, and what does it contain?

The Savarkar Samagra Sahitya is the collected works of V.D. Savarkar in ten volumes, issued by Hindi Sahitya Sadan, gathering the Marathi and English output of sixty years: the poetry and plays, the histories, the autobiography of the Andaman years, the political essays and the Ratnagiri-period writings against superstition. For the study of the 1923 framework the Samagra matters because it preserves the texts in which the author commented on his own definition across four decades, including the anti-superstition essays of the middle 1930s that Anglophone commentary long ignored. Bakhle’s Savarkar and the Making of Hindutva (2024) rests heavily on this Marathi reservoir.

Q: Did Savarkar oppose idol worship?

Yes, as a matter of belief, and the distinction he drew ran between the idol and the institution. In the Ratnagiri-period essays Savarkar treated the attribution of supernatural power to images as superstition and the commerce surrounding pilgrimage as priestcraft, and he demanded that worship justify itself by utility like every other inherited practice. The same man promoted the Patitpavan temple of 1931, because a temple functioned in his design as civic architecture, a hall where the consolidated community assembled across caste, whatever theology the worshippers carried inside. Opposition to idol worship and sponsorship of temples coexisted in his practice without contradiction, because the framework valued the assembly and audited the belief.

Q: In what languages did Savarkar write his books?

Savarkar wrote chiefly in Marathi, with one decisive exception. The Marathi corpus carries the poetry, the plays, the histories, the autobiography Mazi Janmathep, and the first book, the life of the Italian nationalist Joseph Mazzini written in his London period. The exception is the definition itself: Essentials of Hindutva was composed and published in English in 1923, addressed past the Marathi reading public to an all-India and imperial audience, and its English dress is part of its design. Later works moved between the languages, and the ten volumes of the Samagra preserve both streams, which is why serious scholarship on the framework, Bakhle’s included, works bilingually.

Q: What did Savarkar argue in The Indian War of Independence of 1857?

Savarkar wrote the book in London and published it in 1909, and the Government of India proscribed it before copies could circulate. Against the British account of a sepoy mutiny, he argued that the rising of 1857 was a national war of independence, waged by princes, sepoys and peasants together under the twin banners of swadharma, one’s own faith, and swaraj, one’s own rule. The book made his revolutionary reputation, circulated in clandestine editions among Indian students and soldiers for four decades, and supplied the narrative method, a civilisation recovering its nationhood through war, that the 1923 definition would later formalise into criteria.

Q: How was the 1923 book received by Hindu leaders when it appeared?

The reception inside the author’s own camp was immediate and warm. The front matter of the later editions preserves the tributes: Swami Shraddhanand compared the definition’s arrival to a Vedic dawn, Lala Lajpat Rai and Madan Mohan Malaviya hailed the book as an original and scholarly contribution, and C. Vijayaraghavachariar, who had presided over both the Congress and the Hindu Mahasabha, wrote to the author in April 1941 praising the closing chapter in particular. The Hindu Mahasabha converted praise into law by adopting the book’s culminating sentence as its authoritative definition of a Hindu, the institutional act that carried the definition from a prisoner’s pamphlet into a movement’s constitution.