Bhimrao Ramji Ambedkar is known to one kind of reader as the jurist who chaired the drafting of India’s constitution and led the movement against untouchability, and to another kind of reader he is barely a name at all. Both readers tend to assume that his relation to Marx can be settled in a sentence: either he was an enemy of Marxism who buried it, or he was a fellow traveller who borrowed from it. The texts support neither sentence. Ambedkar read Marx directly, at length, and with the care of a man trained to weigh clauses. What he produced was a division of the doctrine into parts he accepted and parts he rejected, with reasons stated for each part, and the division was set out most plainly in a lecture delivered near the end of his life in which he compared Marx with the Buddha point by point. That comparison is the decisive document in the whole debate, and it is read far less often than it is invoked.

The shape of his position can be stated before any of its detail. He accepted the analysis of exploitation, the existence of class conflict, and the charge that private property produces suffering for the many, and he held that these evils must be ended. The refusals were as specific as the acceptances. Dialectical materialism failed as a philosophy; violence failed as a method; the dictatorship of the proletariat failed as an instrument; and the abolition of religion failed as an aim, each for reasons he stated. Against the Indian communist movement of his own time he brought a separate and sharper charge, concerning caste, which this article states in his own terms and at full strength. Against the Marxist replies to all of this, which are also stated here at full strength and attached to the named men who made them, he set an alternative built from constitutional law and state ownership. The conclusion that organises everything that follows is the partial acceptance: a thinker treated by both camps as an opponent of the tradition in fact divided the tradition, accepting its diagnosis and its objective while refusing its method and its philosophy. Recovering that division is more useful than adjudicating a quarrel he never quite had.

Ambedkar and the making of a reader of Marx

Ambedkar was born in 1891 at Mhow, into the Mahar community, which the caste order placed among the untouchables. The circumstances of that birth matter to the argument of this article because they determined what he would look for in any doctrine of liberation: not an abstract promise of equality, but an account of graded hierarchy that named his own people’s degradation and a method that plausibly ended it. His schooling was repeatedly interrupted by the humiliations the caste order arranged for children of his birth, yet patronage and his own ferocious application carried him to Elphinstone College in Bombay and then, on a scholarship from the Maharaja of Baroda, to Columbia University in New York. At Columbia he studied economics, sociology, history, and political science, taking a doctorate for work on provincial finance in British India. In London he added a doctorate in science at the London School of Economics for a study of the rupee, and he was called to the Bar at Gray’s Inn. Few men of his generation, in any country, carried so much formal training into political life.

The training is not decoration. It explains the character of his engagement with Marx, which was never the engagement of a pamphleteer seizing on slogans. He had read political economy as a science before he read it as a creed, and he read Marx the way a drafter reads a bill, clause by clause, asking of each proposition what it asserted, what it assumed, and what it would do if enacted. He taught political economy in Bombay for a period after his return from London, founded newspapers to give his community a public voice, and built one of the largest private libraries in the country at his Bombay house, Rajgruha, because books were the instruments of his politics as much as speeches were. When he sat down with the Marxist corpus, he sat down with the habits of an economist who had audited colonial budgets and a lawyer who had cross-examined witnesses. The result, across three decades of writing, is an engagement that accepts and rejects at the level of the individual proposition, which is exactly why it frustrates partisans of both camps who want a verdict on the whole.

His political career ran alongside the reading. He led the satyagraha at Mahad in 1927 to assert the right of untouchables to draw water from a public tank, directed the temple entry movement at the Kalaram temple in Nashik, and represented the depressed classes at the Round Table Conferences in London, where his demand for separate electorates brought him into the famous confrontation with Gandhi that ended in the Poona Pact. He served as labour member of the Viceroy’s Executive Council during the war years, founded political parties to carry his people’s representation into legislatures, became the first law minister of independent India, and chaired the committee that drafted the constitution, resigning from the cabinet in 1951 when his Hindu Code Bill was stalled. In 1935 he had declared that though he was born a Hindu he would not die one, and in October 1956, at Nagpur, he formally converted to Buddhism together with a vast gathering of his followers. He died in Delhi in December of that same year, a few weeks after delivering the lecture on Marx and the Buddha that this article treats as the key to his position.

How did his economics training shape the engagement?

It made him read doctrine as a set of costed propositions. A man who had audited provincial budgets asked of every Marxist claim what it assumed about property, labour, and the state, and what machinery would be needed to carry it out, rather than whether it sounded righteous.

That habit shows most clearly in the difference between his treatment of Marx and his treatment of Marxists. Toward the doctrine he is a fellow examiner, testing each element against evidence and against the ends it professes. Toward the movement in India he is a prosecutor, because the movement was a going concern with leaders, programmes, and a record, and it could be charged with specific failures in specific years. The two treatments are often confused by readers who import the prosecutor’s tone into the examiner’s pages, or who soften the charge against the movement by pointing to the care of the doctrinal analysis. Keeping the two apart is the first discipline this article observes, and it is a discipline the brief for the Indian cluster requires, because the analytical separation of caste and class belongs to a companion article while the movement’s own history belongs to another. What belongs here is the debate itself: what one rigorously trained mind, speaking for the most oppressed community in his country, accepted from Marx, what he rejected, and what he charged against those who claimed Marx’s authority in India.

There is a further reason the engagement matters beyond biography. Ambedkar was not commenting on an Indian problem from outside Marxism’s concerns, nor was Marxism commenting on a peripheral case. Caste posed, and poses in the texts, a direct challenge to the Marxist account of how societies divide: a hierarchy at once economic and ritual, hereditary and occupational, sanctioned by religion and enforced by custom, which sorted labourers before any labour market could form. A theory of class that could not digest caste would be incomplete on its own terms, and Ambedkar knew it. His criticism therefore carried a weight that criticisms from liberals or conservatives did not. He accepted enough of the Marxist diagnosis to make his rejection of the remainder impossible to dismiss as ignorance, and he spoke from a constituency whose emancipation was the test any Indian doctrine of liberation had to pass. That combination, insider’s knowledge of the texts and outsider’s standing in the movement, is what gives the debate its continuing intellectual charge.

The two texts that carry the debate

Two documents carry nearly the whole of the argument, and a reader who knows them knows the debate. The first is the address he prepared in 1936 for a conference of caste reformers at Lahore, an address that was never delivered because the reformers who had invited him read the text and cancelled the conference rather than let it be spoken. He printed it himself, under the title Annihilation of Caste, and it became the most famous statement of his social philosophy. Its subject is caste rather than Marx, but it contains the analysis on which his later quarrel with Marxism rests: caste divides not labour but labourers, graded, hereditary, and religiously sanctioned, and a hierarchy so constructed prevents the formation of the united working class that Marxist strategy presupposes. The second document is the lecture of November 1956, delivered at Kathmandu during a conference of the World Fellowship of Buddhists and published after his death under the title Buddha or Karl Marx. There the comparison is explicit and itemised: he states what he accepts from Marx, what he rejects, and what alternative he proposes, in the form of a direct confrontation between two teachers separated by more than two thousand years.

Why was the lecture on Marx and the Buddha delivered in Kathmandu?

Because his hearers were the assembled World Fellowship of Buddhists, and he wished to argue before Buddhists themselves that their founder had addressed the problem of suffering and property long before Marx, and had solved it by a method that made violence unnecessary. The setting was the argument.

Between the two documents lies a third, less famous and almost as important: the memorandum he submitted to the Constituent Assembly in 1947 under the title States and Minorities, which set out his economic programme in constitutional form. Where the address supplies the analysis of caste and the lecture supplies the philosophy of ends and means, the memorandum supplies the mechanism: a scheme of state socialism, with key industries owned by the state and agriculture reorganised as a state industry, entrenched in the constitution itself so that no ordinary parliamentary majority could undo it. Any account of Ambedkar and Marx that rests on the lecture alone will miss the fact that his alternative to the Marxist route was not a pious hope but a drafted scheme, drawn up by a man who was simultaneously drafting a constitution. The three texts together, spanning twenty years, show a position that was stable in its architecture: the diagnosis of exploitation accepted throughout, the philosophy and the violent method refused throughout, and the institutional alternative elaborated with increasing precision.

The afterlife of the two famous texts explains much of the confusion the debate attracts. Annihilation of Caste was answered at once, most notably by Gandhi, who reviewed it in his journal with respect for its author and resistance to its conclusions, and Ambedkar’s rejoinder was bound into the second edition, so that the text circulates as a controversy rather than a treatise. Buddha or Karl Marx, delivered weeks before its author’s death, circulated more quietly, in editions of his collected writings, and is still more often cited in fragments than read entire. Each camp in the later quarrel learned to quote the half of the record that suited it: the Marxist claim that he was hostile to the tradition rests on the rejected half of the lecture, and the Ambedkarite claim that he had no use for Marx rests on the same half read from the other side. The accepted half, in which he endorses the analysis of exploitation and the ends of communism in terms, sits in the middle of the lecture where neither slogan reaches. This article proceeds through the record in his own order: first the address and its analysis of caste, then the lecture and its itemised comparison, then the charge against the movement, then the replies, and finally his answer and its acknowledged weakness.

The undelivered address: caste as a division of labourers

The address of 1936 opens against its own audience. The reformers of the Jat-Pat Todak Mandal, a society devoted to breaking caste, had a programme of their own: begin by abolishing the sub-castes, they held, and caste itself would soften in time. Ambedkar told them their starting point was wrong. Sub-castes, he argued, are not a corruption of an otherwise sound arrangement; they are the same principle at a smaller scale, and a reform that attacked the branches while protecting the root would strengthen the root. The root was caste itself, and his address proceeded to define what caste is with a precision that still organises the discussion. “Caste is not a division of labour. It is a division of labourers.” The sentence is the hinge of the text. Division of labour exists in every complex society and offends no principle of justice by itself. Caste divides the labourers themselves into watertight compartments, ranks the compartments in a fixed order of reverence and contempt, and assigns each person to a compartment at birth, so that the work a man does is decided not by training, inclination, or capacity but by the accident of the family into which he is born.

How did caste block the formation of a united working class, in his argument?

By sorting labourers into graded compartments before any workplace could unite them. A hierarchy that taught each group to look down on the group below gave workers an interest in the system that divided them, so solidarity across the whole class could not form on economic ground alone.

The grading is the feature that does the most analytical work. Caste, as he described it, is not a simple two-storey structure of high and low. It is an ascending scale of reverence and a descending scale of contempt, in which almost every caste has another caste beneath it to despise and another above it to envy. A system of that shape is unusually resistant to collective action, because it distributes small privileges of status down the whole ladder and gives nearly everyone a stake in preserving the rung below. The economic consequences follow directly. Occupations are not chosen but inherited, so the labour force cannot move to where it is needed or valued. Men are graded rather than their work, so efficiency and merit have no purchase. Most importantly for the Marx question, the precondition of class politics, a working class that experiences itself as one body with one antagonist, is undermined in advance: the labourers are divided among themselves by a principle older and more intimate than the wage relation, sanctified by religion and reproduced in marriage. A doctrine that expected the Indian proletariat to unite as a class had first to explain what would dissolve the compartments. Ambedkar’s address is, among other things, a demand for that explanation.

The address also identifies the mechanism that reproduces caste across generations: endogamy. Caste is maintained, he argued, because marriage is confined within the group, so that the compartment renews itself in every generation regardless of what happens in the economy. This is why he held inter-caste marriage, and not fusion of sub-castes or temple entry or any reform of manners, to be the real solvent, since it attacks the machinery of reproduction itself. The point matters for the comparison with Marx because it locates the persistence of the hierarchy in a social institution, the family and its marriage rules, rather than in the organisation of production alone. A change in property relations, on his analysis, would leave the marriage market and the kitchen rules intact, and with them the daily machinery by which graded inequality is taught, enforced, and renewed. Whatever else a revolution accomplished, it would not by itself persuade a father to give his daughter across the line. That is the concrete form in which the superstructure, in his analysis, declines to follow the base.

What did the conference organisers refuse to let him say?

The text attacked the authority of the sacred scriptures and declared that caste could not be destroyed while the scriptures that sanctified it were held sacred. The reception committee demanded deletions from those passages; he refused to alter a line, and the conference was cancelled.

The cancellation is instructive because it confirmed the address’s own thesis in the act of suppressing it. The reformers could contemplate breaking caste as a social arrangement; they could not contemplate breaking the religious authority that underwrote it. Ambedkar’s argument was precisely that the second breaking was the precondition of the first. Caste, he held, is not an arrangement Hindus happen to practise; it is an arrangement their scriptures command, and a man who believes the shastras are divine will experience obedience to caste as piety rather than prejudice. Hence his formulation that caste is a monster crossing the path of every reform, political or economic: attempt any change, and at some point the change requires ordinary people to act against what they have been taught is sacred, and the reform dies there. His conclusion was correspondingly severe. The sacredness of the scriptures had to be destroyed, the authority of the priesthood broken, and religion rebuilt on principles of liberty, equality, and fraternity, or caste would survive every programme erected against it. Marxists reading the address should note the structure of the claim. It does not deny that caste serves material interests; it denies that material interest exhausts caste’s defences. The hierarchy is guarded by belief, and belief must be fought on its own ground.

Religion and reform: why the hierarchy is guarded by belief

The address was written for reformers, and its most consequential passages concern the sequence of reform. His hosts wanted social reform to precede and prepare political change; others in the national movement argued that political freedom must come first and social reform would follow. Ambedkar cut across both positions with an observation drawn from the untouchables’ experience: reforms are not carried by manifestos but by the willingness of ordinary families to eat together, draw water together, worship together, and marry across lines, and that willingness is governed by religious conviction. A society may change its rulers and its laws and leave the conviction untouched, in which case the hierarchy reproduces itself under new management. This is the seed from which his whole divergence from Marxism grows. If the decisive obstacles to emancipation are located in belief and custom as much as in property, then a programme that assigns belief and custom to the superstructure, expecting them to wither once the base is transformed, has misread the order of causation in the Indian case. He did not argue that property was unimportant. He argued that in India the ideological was load-bearing.

It is worth pausing on the fairness of this to Marx himself, because Ambedkar was scrupulous about it in his later work. The mature Marxist tradition contained resources for a subtler view: Engels had written of the relative independence of ideological forms, and later Marxist historians would treat caste as a formation with its own weight. Ambedkar’s charge in the address is directed at the practice of reformers and, by anticipation, at any Indian movement that proposed to postpone the caste question until after the seizure of power or the winning of independence. The theoretical refinement belongs to the replies considered later in this article, in the section on the Marxist answer given by named respondents such as Dange and Kosambi. The address itself establishes the fact to be explained: a hierarchy of labourers, graded and sanctified, reproducing itself through marriage, and barring the road to every reform that declines to fight it at the level of belief. The analytical working-out of how caste and class interlock, rather than merely coexist, is the business of the companion study of caste and class in India, and it is deliberately not duplicated here.

One more element of the address deserves record before the narrative moves to the lecture, because it is the element most often omitted in summary. The address does not only attack caste; it sketches the society that should replace it. Hindu society, as he read it, lacked the ideals that make a society a society: liberty, equality, and fraternity were, in his phrase, conspicuously absent from its operative principles, whatever its metaphysics said about the unity of existence. He held fraternity to be the root of the other two, the sentiment that makes a man treat another’s good as his own concern, and he found that sentiment systematically destroyed by a system that graded men at birth. The positive triad matters because it reappears twenty years later as the standard by which he judges both Marx and the Buddha. The lecture’s verdict on Marxism is not that its ends are wrong; it is that its means endanger fraternity, and that without fraternity the ends themselves curdle. A reader who grasps that the same triad governs the address of a young reformer and the lecture of an old convert has grasped the continuity of the position, and can then see the comparison with Marx as what it was: the application, to the strongest doctrine of liberation available, of a standard formed in the struggle against caste.

The Kathmandu lecture: the comparison stated directly

Twenty years separate the address from the lecture, and the later text has a different temper. The address argues with reformers; the lecture instructs an audience of Buddhists in what their tradition should learn, and refuse to learn, from the most successful doctrine of social transformation in the modern world. He begins by setting out the creed he is examining. Stripped of its philosophical apparatus, he says, the teaching of Marx comes to four propositions: that the function of philosophy is to reconstruct the world rather than merely to explain it; that there is a conflict of interest between class and class; that this conflict expresses itself in the exploitation of the have-nots by those who have; and that private property, by concentrating power in the hands of its owners, produces suffering for the many. He then does what his later reputation often denies that he did. He declares his agreement with the substance of these propositions, and says plainly that the ends they name, the ending of exploitation and of the suffering that property inflicts, are ends he shares. The comparison with the Buddha that follows is therefore not a confrontation between a socialist and an anti-socialist. It is a confrontation between two routes to a destination both parties accept, conducted by a man who had spent his life organising the victims of the society in question.

The structure of the lecture is a series of paired examinations. On the diagnosis of suffering, he finds the Buddha and Marx in accord, though the Buddha stated the problem of sorrow and its cause more than two thousand years earlier. On property, he finds the monastic order founded by the Buddha practising common ownership of the requisites of life under rules that bound every member equally, an anticipation of common ownership achieved without confiscation or terror. On method, he finds the decisive difference: the Buddha sought to change the mind of man, by persuasion, by moral teaching, and by the discipline of a voluntary order, while the Marxist route relied on force, on the seizure of power, and on a dictatorship that was supposed to wither and showed no sign of withering. On religion, he finds Marxism committed to treating religion as an opiate and an illusion destined to disappear, while his own conviction was that society cannot do without religion, provided religion is reconstructed as morality in accordance with reason. Each pair is argued, not merely asserted, and the lecture closes by assigning the palm to the Buddha on the ground that ends achieved by persuasion endure, while ends achieved by force must be maintained by force.

What four points did he say he shared with Marx?

He listed them in his own summary of the creed: philosophy exists to reconstruct the world, not merely to interpret it; classes stand in real conflict of interest; the conflict issues in exploitation of those without property; and private ownership concentrates power and manufactures misery for the many.

Precision about the lecture’s occasion sharpens its meaning. He spoke in November 1956, weeks after his conversion at Nagpur and weeks before his death, to an international Buddhist gathering in Kathmandu. He was therefore not performing for a political party or trimming for an election. He was delivering, to the community he had just joined, his settled account of the relationship between its founder’s teaching and the doctrine that dominated the social thought of the century. The lateness matters for the historian: this is a considered testament, not a youthful polemic, and it agrees in structure with the memorandum of 1947 and the address of 1936 even where it differs in vocabulary. It also matters that the lecture is comparative in form rather than denunciatory. He does not say that Marx was wrong about the world; he says that the Buddha was right earlier, and right by a safer road. Readers who approach the text expecting either embrace or anathema will be disappointed by its patience, and the disappointment is the measure of how badly the later quarrel has served the record. The debate, properly so called, begins only when the itemised form of the lecture is respected, which is why the next sections take his four acceptances and four rejections in turn, before turning to the separate charge against the movement and the replies it provoked.

What he accepted: the diagnosis and the objective

The accepted half of the lecture deserves to be taken as slowly as the rejected half, because it is the half that both later camps found inconvenient. Begin with the first proposition, the redefinition of philosophy itself. Marx had written in his youth that philosophers had only interpreted the world, in various ways, when the point was to change it. Ambedkar treated this as the founding gesture of a serious social philosophy and adopted it as his own. His life’s work, from the newspapers he founded to the constitution he drafted, was reconstruction rather than interpretation, and he recognised in Marx’s sentence the charter of that work. The acceptance is not trivial: many of Marx’s critics begin by denying that the world needs reconstruction on Marxist lines, whereas Ambedkar begins by endorsing the imperative and quarrelling only with the blueprint.

What did Ambedkar accept from Marx?

that philosophy should reconstruct the world rather than interpret it, that exploitation exists, that private property produces suffering for many, and that these must be ended. He rejected dialectical materialism, the necessity of violence, the dictatorship of the proletariat, and the abolition of religion.

The second and third acceptances concern class and exploitation. He accepted that society is divided into classes with opposed interests, and that the opposition is not a misunderstanding to be dissolved by goodwill but a structural fact that issues in exploitation. This was not a concession wrenched from him in debate; it was continuous with everything he had written about India. His analysis of caste is itself an analysis of exploitation, indeed of a more thorough exploitation than the wage relation alone describes, since the untouchable was exploited not only as a labourer but as a person, denied water, schooling, worship, and the ordinary dignities of movement. When he read in Marx that the history of class society is the history of exploitation, he was reading a general proposition of which his own community’s condition was the most extreme Indian instance. He did qualify the theory of class in one respect, and the qualification is taken up below: in India, class does not present itself in the clean form the European literature describes, because caste cuts across it and binds each fragment to the hierarchy. But the qualification amends the map; it does not deny the territory.

How did he treat the theory of class?

As true in substance and incomplete in India. Classes and their opposed interests were real, and exploitation was their consequence; yet caste overlaid the class structure with a graded order of its own, so any politics that addressed class while ignoring caste addressed only part of the machine.

The fourth acceptance, private property, is the one that most embarrasses the portrait of Ambedkar as a safe constitutionalist remote from socialist conclusions. He held that ownership of productive wealth by private hands was a source of exploitation and of the suffering property inflicts, and that it must be ended by common ownership rather than merely softened. His own economic proposals, examined in a later section, went to lengths that startle readers who know him only as a drafter of fundamental rights: nationalisation of key industries, state ownership of agricultural land, collective farming under state direction, a state monopoly of insurance. The difference from Marxism lay not in the destination of common ownership but in the vehicle: ownership was to be transferred to the state by constitutional command, and held there under a democratic order, rather than seized by a revolutionary party and held under its dictatorship. A reader who attends only to the ends will find Ambedkar well inside the socialist perimeter. A reader who attends only to the means will find him outside the Marxist one. Both readers are reading the lecture correctly, in halves.

Taken together, the four acceptances define what this article calls the partial acceptance, the claim around which the whole debate should be organised. Ambedkar’s position was not a rejection of Marx but a division of Marx: the diagnosis of exploitation and the objective of ending it stand on one side, the philosophy and the method on the other, and the division is principled rather than tactical. He did not accept the diagnosis in public and disown it in private, nor accept it for India while denying it elsewhere. He accepted it as the correct account of class society and then argued, at the height of his powers and in full view of the consequences, that the Marxist method for curing the disease would reproduce the disease in political form. The two misrepresentations named in the brief for this article both die on this point. The Marxist claim that he was an enemy of the tradition cannot survive his endorsement of the creed’s four propositions. The claim from the other camp, that Marx was of no use to him, dies on the same page. What survives is a debate about means, conducted between parties who shared the ends, and debates about means are the debates in which reasons, rather than allegiances, decide.

The accepted and rejected ledger

The itemised form of his position invites an instrument, and the series supplies one: the accepted and rejected ledger. The ledger sets out each element of the Marxist doctrine in turn, records his position on it, gives his stated reason in compressed form, and names the text in which the position is stated. It is reproduced here because the debate is chronically distorted by summary, and a ledger resists summary; a reader can check each row against the source named in its final column and find the position exactly where the row says it is.

Element of Marxism	His position	His stated reason	Source
The analysis of exploitation	Accepted	Class society rests on the extraction of labour from those who own nothing, and India exhibits the fact in its sharpest form	Lecture comparing Marx with the Buddha
The theory of class	Accepted with qualification	Classes and their opposed interests are real, but caste overlays class in India and divides the labourers themselves	The undelivered address on caste
Private property	Accepted as an evil to be ended	Ownership of productive wealth concentrates power and manufactures suffering for the many who own nothing	Lecture; memorandum on states and minorities
Dialectical materialism	Rejected	A philosophy that explains everything by matter and motion cannot ground the moral transformation society requires	The Kathmandu lecture
Revolutionary violence	Rejected	What force wins must be held by force, and coercion corrupts the order it founds	The comparison with the Buddha
Dictatorship of the proletariat	Rejected	Power seized in the name of a class shows no tendency to surrender itself, whatever the theory promises	Lecture; debates on constitutional method
The treatment of religion	Rejected	Society cannot do without religion reconstructed as morality; to abolish it is to remove the cement of conduct	Lecture; writings on religion and dhamma

Three observations about the ledger repay attention. The first is that the rows are not arranged by importance but by the order of his own exposition, and the acceptances cluster where the doctrine is descriptive while the rejections cluster where it becomes prescriptive. He trusts Marx as an anatomist of class society and distrusts him as an engineer of its successor. The second observation is that every row is sourced, and the sources are his major texts rather than reported remarks, which is what distinguishes the ledger from the folklore of the debate. The third is that the ledger is incomplete in one deliberate respect: it records the doctrine, not the movement. His charge against the Indian communists, which concerns leaders, programmes, and conduct rather than propositions, has no row here, because it belongs to a different order of criticism, and confusing the two orders is the commonest error in the literature. The ledger’s discipline is the article’s discipline: doctrine first, movement second, replies third, and the author’s own answer last.

How does the ledger clarify what summaries blur?

By forcing each element of the doctrine to carry its own verdict. A summary must choose a single adjective for the whole relationship, hostile or indebted; the ledger shows acceptances and rejections side by side, each with its reason and its source, so the pattern of division becomes visible at a glance.

Private property and the ending of suffering

Because the acceptance of the property critique is so often understated, it deserves a section of its own. Ambedkar’s route to the question was not the factory but the village. The Indian village, in his unsparing description, was the hierarchy’s home address: a settlement where caste allocated land, water, and labour by birth and where the untouchable’s relation to property was defined by exclusion: he could not own in any secure sense, could not draw from the common well, could not enter the common temple, and owed labour to the village establishment under hereditary obligation. Property in such a setting was not an abstraction in an economics text; it was the material form of the hierarchy itself. When he concluded that private property in the means of production must give way to common ownership, he was generalising from a scene in which property and caste were two names for a single allocation of power. His agreement with Marx on this point was therefore independent in origin: two analyses, one beginning from the English mill and one from the Indian village, converging on the same institution.

The convergence extended to the remedy in its economic outline. In the memorandum of 1947 he proposed that the constitution itself declare a scheme of state socialism: key and basic industries to be owned and run by the state, insurance to be a state monopoly, and agriculture to be organised as a state industry, with land owned by the state and cultivated by village communities under state direction and support. He further proposed that this economic structure be placed beyond the reach of ordinary legislation, embedded in the fundamental law so that a parliamentary majority could neither install a private order nor dismantle the public one at pleasure. The reasoning is characteristically his. Parliamentary socialism, in his judgement, was an insecure possession, granted by one majority and revocable by the next; a socialism that deserved the name had to be constitutional, as much a part of the state’s permanent frame as the franchise or the courts. Readers of the Indian constitution will search it in vain for this scheme, since the Assembly declined to adopt it, but its existence as a drafted proposal from the chairman of the drafting committee is the standing refutation of any account that makes him a liberal individualist in economics.

What did the memorandum on states and minorities propose for land and industry?

State ownership at both points. Land was to belong to the state and be farmed collectively by village communities with state finance, while key and basic industries, and insurance with them, were to be owned and operated by the state under constitutional command.

The property row of the ledger therefore conceals a subtlety that the later sections need. Ambedkar accepted common ownership as the end and state ownership as its form, but he rejected the sequence by which Marxism proposed to reach it: expropriation by a victorious class acting through a dictatorship, followed at some indefinite date by the withering of the state. In his scheme the state does not wither; it is the permanent trustee of common ownership, disciplined by elections, courts, and a constitution that fixes its economic duties. This is not a detail of administration. It is the institutional form of his whole objection, the belief that power must be caged in law at every moment, including the moment of emancipation, and it explains why the dictatorship of the proletariat, rather than any economic doctrine, is the deepest of his rejections. The concept he refused is examined in its own right in the series study of the dictatorship of the proletariat, and only its place in his division of Marx is treated here. What matters for the ledger is the consistency: the man who wanted property socialised by constitutional clause was bound to refuse a method that socialised it by decree of a committee, because the difference between clause and decree was, for him, the difference between emancipation and a new mastery.

The first rejection: dialectical materialism as a philosophy

His rejection of dialectical materialism was philosophical in the strict sense: he did not believe the doctrine was true. In the lecture he examined the claim that matter in motion, governed by dialectical laws, exhausts reality and explains thought as its product, and he found in it a dogma rather than a discovery. His objections were of three kinds. The first was that the doctrine, as held by its adherents, had ceased to behave like a science and had begun to behave like the religions it despised: its propositions were defended as a creed, its texts quoted as scripture, and deviation treated as heresy, so that a philosophy announcing itself as the enemy of dogma had acquired a priesthood of its own. The second objection was that the economic interpretation of history, in the strong form which made material interest the sole engine of human action, was not borne out by the evidence of his own field. Men in India starved rather than break a caste rule about food, died rather than draw water from a forbidden source, and ruined themselves for ceremonies whose cost no calculus of interest could justify. Whatever moved those men was not reducible to their material interest without emptying the word interest of meaning. The third objection was the deepest: a philosophy that treats mind and morals as reflections of matter cannot explain why anyone should sacrifice himself for a doctrine, nor ground the obligations that a free society needs its members to feel. Emancipation, on his account, is a moral achievement before it is an economic one, and a philosophy with no room for moral reality had disqualified itself from guiding it.

It is important to the record that this rejection coexisted with respect for Marx’s historical writing. He did not deny that material conditions shape societies, nor that the economic interpretation illuminates long stretches of history; he denied its sufficiency and its metaphysical completion. His own alternative, drawn from the teaching of the Buddha as he reconstructed it, held the world to be in flux, relations to be conditioned, and mind to be a real power in the causal order, so that changing men’s minds was not a distraction from changing the world but one of the ways the world is changed. The difference sounds scholastic and is anything but. If mind is epiphenomenal, then persuasion is decoration and force is the only real lever, which is the practical premise of the Marxist method. If mind is causal, persuasion becomes a genuine instrument of transformation, slower than force and more durable. The whole quarrel about means, which occupies the next two sections, is contained in embryo in this quarrel about philosophy, and Ambedkar, who understood the architecture of doctrines, placed the philosophical rejection first in his own order of exposition.

Why did he call the materialist creed a dogma?

Because its adherents, as he observed them, held its propositions the way believers hold a revelation: the texts were treated as final, the founder’s authority settled disputes, and doubt was answered as disloyalty. A doctrine that began as criticism of religion had reproduced religion’s habits of defence.

The second rejection: the necessity of violence

The rejection of violence follows from the philosophy and stands on two arguments, one moral and one practical, which he stated with equal emphasis. The moral argument is that force treats men as material to be shaped rather than as minds to be convinced, and thereby violates the fraternity that the new society is supposed to embody; a liberation achieved by terrorising opponents teaches the liberated that terror is the grammar of politics, and the lesson outlives the emergency that was supposed to excuse it. The practical argument is that violence does not finish its work. An order founded on confiscation and the suppression of the dispossessed must maintain an apparatus of coercion to defend itself against the resentment it has created, so the emergency instrument becomes a permanent institution and the promised dissolution of the state recedes. Against this he set the method he attributed to the Buddha: persuasion, the conversion of the opponent’s understanding, the slow conquest of the mind by teaching and example. What persuasion builds does not need a garrison, because the converted defend the new order as their own conviction rather than endure it as their defeat. He was not a pacifist in the Gandhian mould, and his own movements used the full repertoire of constitutional agitation, boycott, satyagraha, and litigation; what he refused was the doctrine that bloodshed is a necessary midwife of the new society, and the claim that history had licensed it.

His own practice supplies the evidence he would have cited. The Mahad satyagraha asserted a civic right at a public tank and accepted the discipline of a court verdict when the right was litigated; the temple entry campaign pressed a religious claim by presence and endurance rather than by assault; his parties contested elections, his memoranda argued clauses, his newspapers made cases. Even his angriest texts are arguments, constructed to convince a reader who is free to refuse them. There is a straight line from that practice to the lecture’s verdict: the man had spent forty years attempting to move the most rigid hierarchy in the world by reason, organisation, and law, and while the results were incomplete they were not nothing; untouchability stood abolished in the constitution he helped to write, its practice made punishable by law, without a shot fired by his movement. The Marxist answer, that such gains are concessions the propertied order grants to avoid worse and withdraws when it can, is stated in its place below, among the replies of the named respondents. For the structure of his position it is enough to see that the rejection of violence was not squeamishness. It was the methodological commitment of a lifetime, defended philosophically as the only method whose results do not require their own negation.

What did he propose in place of revolutionary force?

Persuasion organised by law: teaching that changes what men believe, organisation that changes what they can do together, and constitutional instruments that change what the state must guarantee, so that the new order rests on conviction and right rather than on a garrison.

A further strand of the argument deserves separation, because it is the strand his Marxist respondents found hardest to answer. Violence, he held, is not only a means but a selection mechanism: it promotes to leadership the men most skilled in conspiracy, discipline, and force, and those men, having won by those arts, govern by them. A movement that proposes to use violence temporarily must therefore explain how the temporary users will be prevented from becoming permanent rulers, and the Marxist tradition’s explanation, that the state would wither once classes disappeared, seemed to him a prophecy doing the work of a mechanism. The Buddha’s method selected differently: it promoted teachers, and the order it founded was governed by the seniority of ordination and the consensus of the assembly rather than by command. Whether or not the contrast is fair to every revolutionary episode, it identifies the question that any doctrine of violent transition must answer, and Ambedkar’s judgement was that Marxism had not answered it but had merely scheduled the answer for a future that its own institutions made unlikely. The point returns in the next section, where the instrument that was supposed to be temporary has a name.

The third rejection: the dictatorship of the proletariat

The dictatorship of the proletariat was, for Ambedkar, the point at which the Marxist method disclosed its destination. His treatment of it in the lecture is brief and deadly. The doctrine asks men to accept a dictatorship, the rule of one class organised as a state without legal restraint, on the assurance that it is transitional: when classes have been abolished, the state will wither away and government of men will give place to administration of things. He replied that he could see no reason to believe the assurance, and several reasons to doubt it. Dictatorships, in the record available to him, did not dissolve themselves; they consolidated. The men who hold unconstrained power acquire interests in holding it, institutions grow around the holding, and the emergency that justified the concentration is renewed as required. A theory that makes the surrender of power depend on the virtue of those who hold it has, for a constitutional lawyer, simply failed to specify its most important clause. He had spent the same years drafting a constitution whose entire art consisted in not trusting power: dividing it, terming it, reviewing it, making it answerable. The dictatorship asked him to trust power absolutely, in the name of the class whose suffering he had spent his life documenting, and he refused the request in terms that allowed no misunderstanding.

His refusal was not a defence of the existing order, a point the fairness of the record requires. In his final address to the Constituent Assembly he warned that political democracy, one man one vote, could not endure on a foundation of social and economic inequality: “political democracy cannot last unless there lies at its base social democracy”. The sentence states his double position exactly. He wanted the social revolution; he distrusted both the gradualism that postponed it forever and the dictatorship that promised it at the price of the political revolution already won. Parliamentary democracy, for him, was not the terminus but the instrument, valuable precisely because it allowed the social question to be pressed by those who suffered from it, in numbers, through organisations and elections, without surrendering the weapon to a guardian class. The dictatorship inverted the arrangement: it disarmed the sufferers in order to save them, and required them to believe that the saviours would return the weapon voluntarily. The series treats the concept of the dictatorship in its own study, linked earlier in this article; the concern here is narrower, with what its rejection reveals about the architecture of his thought. Every other rejection in the ledger can be derived from this one. Violence is refused because it installs the dictatorship’s personnel; materialism is refused because it supplies the dictatorship’s excuse, that men are matter to be arranged; religion is defended because a moral order independent of the state is one of the few checks on a power that claims to embody history.

Against the obvious rejoinder, that the dictatorship is by definition the rule of the vast majority over a dispossessed minority and therefore democratic in substance, his answer was institutional rather than rhetorical. Majority rule earns the name of democracy only when it is exercised under forms that protect the minority’s right to become a majority: free elections at fixed terms, an independent judiciary, a press that cannot be licensed into silence, and a constitution that even the majority cannot amend at pleasure. Strip those forms away and majority rule is a census followed by a command. He had direct experience of what majorities do to minorities when unconstrained, since the community he led lived at the mercy of a permanent majority in every village in India, and his political life was a long effort to build constitutional fences against exactly that condition. A doctrine that proposed to cure majority tyranny by removing the fences, on the ground that the majority would now be the right majority, asked the victims of the old fences’ absence to applaud their demolition. He declined, and the decline is the most characteristic act of his political intelligence.

The fourth rejection: the abolition of religion

The treatment of religion is the rejection that surprises readers formed by the usual portraits, since the scourge of the Hindu scriptures might be expected to welcome a doctrine that dismissed religion altogether. He did the opposite. Marxism, as he presented it, regarded religion as an illusion characteristic of man’s immaturity, an opiate that would disappear when the conditions that required it disappeared. Ambedkar replied that society stands in permanent need of religion, rightly understood, and that the task is not abolition but reconstruction. His distinction, elaborated across his writings on religion and dhamma, separates two things that share a name. Religion as he found it in the caste order meant ritual, priestcraft, scriptural sanction, and rules governing eating, marrying, and touching: a machinery for assigning men their places. Dhamma, the word he preferred for what the Buddha taught, meant morality: the principles by which men govern their relations with one another, tested by reason and summarised in liberty, equality, and fraternity. The first kind of religion he attacked all his life with a ferocity no Marxist matched. The second kind he held to be indispensable, because a society’s members must be bound by something they regard as sacred, or law and interest alone will not hold them to the duties that freedom leaves unpoliced.

What did he mean by religion, and what did he oppose?

He opposed religion as ritual command and scriptural sanction, the form that assigns men their places at birth. What he defended was religion reconstructed as morality, a sacred order of conduct tested by reason, under which men are bound to treat one another as equals.

The argument cuts in two directions, and both cuts matter for the debate. Against the Hindu order, it denies that caste can claim the protection owed to religion, since a system that degrades men fails the moral test by which genuine religion is known; this is the argument of the address of 1936 pursued into his final writings. Against Marxism, it denies that the moral function can be safely left to wither. If men cease to regard anything as sacred, he reasoned, they will not become rational administrators of things; they will become competitors without restraint, and the strong will call the result nature. A communist society that abolished religion before it had formed the moral habits of communism would therefore destroy its own foundation in the act of clearing the site. There is an irony he did not miss: the tradition that called religion the opium of the people had produced states that demanded of their citizens a faith, in the party, the plan, and the future, more absolute than most churches require. The series treatment of Marx on religion explained sets out the doctrine he was answering; his answer, in compressed form, is that religion is not one thing, that its caste form deserves destruction, and that its moral form is the precondition of the free society both he and Marx professed to want. It is the rejection in which the distance between the two men is greatest and the agreement about ends is most visible, since the argument is conducted entirely within the question of what human beings need in order to live as equals.

His conversion of 1956 is the practical corollary of this section, and it should be read as an act of social philosophy rather than a private consolation. When he led his followers out of the Hindu fold and into Buddhism, he was enacting the reconstruction the lecture describes: exchanging a religion of birth and sanction for one he had reinterpreted as a rational morality, whose scriptures he then rewrote in a single volume for his people’s use. The Buddha and His Dhamma, published after his death, presents the founder not as a god or a metaphysician but as a social teacher whose doctrine answers the problem of suffering with a programme of conduct. Whatever judgement theology passes on that reconstruction, its place in the debate with Marx is clear. It is the working model of the alternative method: not the abolition of religion in favour of a materialist creed, but the replacement of a degrading religion by an emancipating one, carried by persuasion and embraced by an act of collective will. To call the conversion a retreat from social analysis, as some summaries do, is to mistake the instrument for the abandonment of the work.

The Buddha as the alternative route

The lecture’s positive proposal can now be stated whole. The Buddha, in Ambedkar’s reading, had confronted the same facts that Marx confronted: a world of suffering produced in large part by property and by the antagonisms property breeds. His response differed at the point of method. Where the Marxist route proposed to seize the state, expropriate the owners, and govern through a transitional dictatorship until men were remade, the Buddha proposed to remake men directly, by a teaching that altered what they desired and how they regarded one another, and he built an institution, the sangha, in which the new relations could be practised at once. The sangha’s rules, as Ambedkar dwelt on them, are the empirical core of the comparison. The monks held no private property; the requisites of life, robes, alms food, lodging, were held under rules of common use; decisions were taken in assembly by procedures that prized consensus; and admission was open without regard to birth, so that the order embodied within itself the castelessness that the surrounding society denied. Here, in his presentation, was common ownership without confiscation, equality without a police apparatus, and communism of a sort, achieved by consent twenty-five centuries before it was attempted by force. The anticipation matters less as a claim of priority than as an existence proof: the thing Marxism said required dictatorship had existed, in miniature and under discipline, without it.

Where did he place the sangha in the comparison?

At its centre, as evidence. The monastic order showed common ownership, equality of members, and government by assembly operating together without confiscation or coercion, which allowed him to argue that the ends of communism did not logically require the means of dictatorship.

From the sangha he generalised the method of persuasion. Social transformation, on his account, proceeds through three labours that force cannot perform. Teaching must change what men believe about one another, since a hierarchy believed to be sacred will be obeyed under any property regime. Organisation must change what the oppressed can do together, since conviction without combination is a sermon. Law must change what the state guarantees, since gains that rest on goodwill are revocable at goodwill’s pleasure. The Buddha supplied the first labour and modelled the second; the constitution, in Ambedkar’s own century, supplied the third. It is not accidental that the lecture ends in exhortation rather than analysis: he was recruiting his audience to a method, and the method was the one his own life had practised. Nor should the optimism of the proposal be overstated by the reader. He knew persuasion’s pace. He had watched a tank satyagraha litigated for a decade and a temple campaign exhaust itself against a closed door. His claim was never that persuasion is swift; it was that what persuasion wins stays won, because it is held by the minds that were changed, while what force wins must be guarded forever against the minds that were not. Permanence, in the lecture’s closing argument, is the true measure of a method, and by that measure the slower road is the shorter one.

This is the point at which the partial acceptance becomes a programme rather than a verdict, and the place to note what kind of socialism emerges from it. The ends are socialist: common ownership of productive wealth, the ending of exploitation, the abolition of the power that property confers. The means are constitutional and pedagogic: a state bound by fundamental law, an economy socialised by clause rather than by decree, a people educated into the morality the new order requires. The philosophy is neither materialist nor idealist in the textbook senses: it treats mind as a real force and morality as a social technology, and it assigns religion, reconstructed, the task of carrying the morality that law cannot reach. Whether the programme can work is the question the replies press, and it is taken up after the charge against the movement has been heard, because the charge explains why the programme was addressed to India in the form it took. A doctrine of persuasion and law was being proposed to a country in which the organised Marxist movement, by his account, had already failed the oppressed in a specific and documented way.

The charge against the communist movement

Everything so far has concerned doctrine. The charge that follows concerns men and organisations, and it must be stated as he stated it: at full strength, without softening, and without any defence of the movement interposed, because the discipline of this cluster requires his criticism of the communists to stand in his own terms before any reply is heard. The charge has four counts, and they accumulate. The first count is composition. The leadership of the Indian communist movement, he charged, was drawn overwhelmingly from the privileged castes, from the twice-born strata whose members had never known untouchability from the inside and whose social instincts had been formed by the hierarchy the movement claimed to oppose. A party so led carried caste within itself, in its habits of deference and command, whatever its programme said. The second count is programme. The movement’s doctrine, as preached and practised in India, treated caste as a secondary matter, a feature of the superstructure that would dissolve of itself once property relations were transformed, and it therefore had nothing to say to the untouchable about untouchability itself: not about the well he could not draw from, the temple he could not enter, the road he could not walk, or the marriage that could not be made. On the questions that constituted his daily humiliation, the programme was silent, and the silence was not an oversight but a consequence of the theory as the leadership applied it.

The third count is conduct. Where the movement organised workers, he charged, it organised them as workers in the abstract and left the graded hierarchy among them untouched, accepting the caste composition of the workforce and the caste allocation of tasks as facts of the trade rather than as injustices to be fought within the union itself. Strikes were called for wages and against employers; they were not called against the foreman who would not drink from the same vessel or the custom that reserved the clean work for the clean castes. The movement thus recruited the labour of the oppressed for its battles while declining to fight the battle that was theirs, and it did so, on his reading, because to fight that battle would have cost it the allegiance of the caste-Hindu workers it also courted. The fourth count is trust, and it is the conclusion the others support. A movement composed as the first count describes, silent as the second describes, and behaving as the third describes, could not be entrusted with the annihilation of caste. If such a movement came to power, he held, its victory would reproduce the hierarchy under new management: the twice-born would administer the revolution as they had administered everything else, and the untouchable would exchange an old master for a new one who spoke the language of his emancipation. He therefore warned his people against the movement in terms that admitted no ambiguity, and he treated its professions of solidarity as a claim to be tested against its personnel and its record, on which test it failed.

The charge must also be located in its evidence, because he did not make it from a distance. He had watched the movement’s conduct in the mill districts and in the legislatures, had measured its manifestos against the caste question year by year, and had himself been the object of its polemics when he insisted that social reform, the annihilation of caste, could not be postponed until after the political revolution the movement promised. The postponement was, for him, the heart of the matter. Every Indian movement of emancipation he had known had asked the untouchable to wait: the national movement asked him to wait for independence, the reformers asked him to wait for the softening of custom, and the communists asked him to wait for the revolution, after which caste would supposedly wither with the state. His whole political life was a refusal of that sequence. A hierarchy that reproduced itself daily, in water and food and marriage, would not wither while its victims waited for a transformation elsewhere; it had to be fought where it lived, first and continuously, and a movement that declined to fight it there had disclosed its priorities. The charge, in this fullest form, is not that the communists were insincere. It is that their theory, as they applied it, instructed them to be silent at the exact point where his people’s oppression was most itself, and that men formed by the privileged castes found the instruction easy to obey. The replies that the Marxist tradition made to arguments of this kind are given in their own section below, attributed to the named respondents who developed them. They are not given here, and nothing in this section should be read as anticipating or answering them, since the charge deserves the same courtesy he demanded for the doctrine: to be heard entire before it is judged.

One boundary of the charge should be marked, since he marked it himself. The indictment is of the Indian movement of his time, its leadership, programme, and conduct, and it is framed against the background of a doctrine whose diagnosis he had already accepted in large part. He did not argue that Marxists everywhere and always must be indifferent to a hierarchy like caste; he argued that these Marxists, in this country, formed by these castes and armed with this reading of the theory, had been so in fact. The distinction preserves the coherence of the partial acceptance: a man may accept a diagnosis and prosecute its local practitioners for malpractice, and the prosecution is sharper, not weaker, for the acceptance behind it. It also sets the question the replies had to answer. They could not rest on the doctrine’s intentions. They had to show, from the movement’s own materials and from the theory’s resources, that caste had been theorised and fought as more than a shadow of property. Whether the named respondents achieved that showing is the subject of the next sections.

Doctrine and movement: the separation the debate requires

The preceding sections have observed a separation that is worth making explicit, because the failure to observe it is the commonest defect in writing on this subject. Criticism of a doctrine addresses propositions: whether exploitation is real, whether materialism is true, whether dictatorship can be transitional. Criticism of a movement addresses a historical organisation: who led it, what its programme contained, what it did in particular strikes and elections and years. The two criticisms have different evidence, different standards, and different remedies, and an argument that migrates between them without signalling the move will corrupt both. Ambedkar himself kept them apart with unusual consistency. His doctrinal criticism is conducted against texts, chiefly in the lecture, and is answered, where it is answered, by other texts. His movement criticism is conducted against a record, in speeches and polemics across three decades, and is answered, where it is answered, by counter-records and by theoretical work claiming that the movement’s theory, rightly read, never licensed the indifference he charged. The replies considered next belong mostly to this second kind, though the strongest of them, the objection about mechanism, reaches back across the separation and strikes at the doctrine of his alternative rather than at his reading of Marx.

How does the charge against the leadership differ from an argument about theory?

It is answerable by evidence about persons and conduct rather than by exegesis. One could concede every word of it and still defend the doctrine, or refute it entirely and still reject the doctrine, since a movement’s failures do not by themselves settle what its founding texts assert.

The separation also disciplines the reader’s judgement about fairness. The doctrinal debate can be assessed on the lecture alone, and the lecture is a model of the examiner’s courtesy: the creed is summarised in terms its adherents would recognise, the agreements are declared before the disagreements, and the alternative is proposed in the open. The movement debate cannot be assessed on any single text, because it concerns a record spread over decades and districts, and both the charge and the counter-record are necessarily selective. What can be assessed is the structure of the exchange: whether the charge identifies a real obligation the movement bore, and whether the replies meet the charge at the point where it is made. The obligation is not in dispute between serious parties. A movement claiming to emancipate labour in India bore some obligation to say what would become of caste, and the replies are best understood as three ways of discharging that obligation late: by theorising caste as a form of class relation with deep roots in production, as Dange attempted; by Kosambi’s formula of caste as class under primitive conditions of production; and by incorporating the fight against caste oppression into the programme of democratic revolution, defended by the movement’s record of organising across caste, as Namboodiripad argued. Each reply is now taken in turn, at the strength its author gave it.

The Marxist reply, stated in its own terms

The replies deserve the same discipline as the charge: full strength, attributed to named respondents, in their own terms. Three lines of reply were developed within the Indian Marxist tradition, and a fourth, the objection from mechanism, cuts across all of them. The first is associated with S. A. Dange, a founding figure of the movement, whose study of ancient India published in 1949 attempted to give caste a materialist genealogy. Caste, on Dange’s account, was not an eternal religious mystery but a historical product: it arose as primitive communal society dissolved into classes, and its divisions hardened around the organisation of production in a stagnating slave and early feudal order, with the priesthood sanctifying after the fact an arrangement that served the propertied. If caste was made by a mode of production, it could be unmade by one; the fight against caste was therefore not a distraction from the class struggle but a chapter of it, and the movement’s silence, where it existed, was a failure of application rather than a consequence of the theory. The reply concedes the terrain of the charge, the movement’s record, while contesting its explanation: the fault lay in the movement’s immaturity, not in Marxism, and the remedy was a better Marxism rather than its abandonment.

How did Dange answer the caste question from within Marxism?

By giving caste a material birth. In his history of ancient India, caste emerged as communal society broke into classes and production stagnated, and religious sanction followed the arrangement it served; a hierarchy made by production could therefore be unmade by transforming production.

The second reply is the most theoretically ambitious, and it belongs to D. D. Kosambi, the mathematician and historian whose work remade the Marxist study of ancient India. Kosambi described caste as class at a primitive level of production: in a society where the techniques of production were too undeveloped to support a market in labour, the division of labour was frozen into hereditary groups, and caste performed the functions that class performs elsewhere, allocating work, extracting surplus, and fixing status, with the peculiar Indian feature that the freezing was sanctified by religion. The elegance of the formula is that it neither dismisses caste as an illusion nor treats it as an autonomous mystery. Caste is class in a specific historical form, and its extraordinary durability is explained by the extraordinary durability of the village economy that sustained it: where production barely changes for centuries, the social form that organises production barely changes either. On this account the remedy follows the diagnosis with unusual precision. Develop the productive forces, dissolve the self-sufficient village, draw labour into common enterprises, and the material basis of caste erodes; the movement’s task is to hasten that erosion by organising the labourers whom the new economy assembles, across the lines the old economy drew. Kosambi was no apologist for the movement’s practice, and his history is unsparing about priestcraft and kingship alike, which gave his defence of the theory a credibility that party statements lacked.

What did Kosambi make of caste and class?

He treated it as class in frozen form. Where production remained at a primitive level and labour could not be bought and sold freely, the division of labour hardened into hereditary caste, so caste performed the work of class under the cover of religion.

The third reply is institutional and is associated with E. M. S. Namboodiripad, the movement’s most substantial theorist of the caste question in practice, whose writings on Kerala traced the intertwining of caste and landholding in his own region across centuries. Namboodiripad’s position developed over his long career in a direction the charge had demanded: the fight against caste oppression came to be stated, in his work and in the programme his party adopted, as a distinct and necessary struggle within the broader democratic movement, not as a consequence that could be left to follow economic change automatically. The movement’s record in Kerala was offered as evidence: peasant associations and unions of agricultural labourers had organised across caste lines around tenancy, wages, and land, and the social reforms that followed communist-led governments there, in land relations and education, had struck at the material supports of the hierarchy. On this reply, the caste-blind organising that Ambedkar condemned was a misdescription: organising labourers as labourers was itself an attack on caste in its material root, since the caste order rested on the hereditary allocation of labour, and every union that combined the castes in a common demand rehearsed the equality the doctrine promised. The reply also had its defensive edge. The movement’s leaders were drawn from the educated strata of a society in which education had been a caste monopoly; to treat their birth as disqualifying was, the respondents argued, to reproduce the logic of birth that the annihilation of caste existed to destroy, and the test of a leader was his programme and his practice, not his parentage.

How did the Kerala reply handle the charge of indifference?

By pointing to the record of organisation. Peasant and labour unions that combined all castes around land, tenancy, and wages were presented as the annihilation of caste in its material form, and later programme statements named caste oppression as an evil to be fought directly.

The fourth element of the reply is not about caste at all, and it is the strongest objection this article is required to state. It concerns means, and it was pressed by the tradition’s polemicists, in the line of argument associated with B. T. Ranadive’s polemics against reformist routes, against every programme that promised emancipation without power. Stated at full strength it runs as follows. Ambedkar accepts the end: exploitation must end, private ownership of productive wealth must yield to common ownership, the suffering property causes must be abolished. He rejects the only means by which propertied interests have ever been dispossessed, the seizure and use of state power by the dispossessed themselves, and substitutes persuasion, moral transformation, and constitutional enactment. But propertied interests do not surrender their property because they have been persuaded it is unjust, and a parliament elected under conditions of property will reflect the power property buys: in the press, in the parties, in the courts, in the administration. A constitution commanding socialism will be interpreted by judges drawn from the propertied world, administered by officials formed in it, and amended or evaded when it bites. The objection to the means therefore leaves the objective unattainable: the programme is a destination drawn on a map with no road, and its author’s horror of the road that exists is a moral preference, not a strategy. This objection does not depend on any claim about caste. It strikes at the joint where Ambedkar’s alternative bears its weight, and his answer to it, given in the next section, is the most instructive part of the entire debate, both for what it builds and for the weakness it cannot quite conceal.

His answer: constitutional instruments and state ownership

Ambedkar’s reply to the mechanism objection was not improvised; it was the memorandum of 1947 read as a strategy. Its core move is to relocate the transfer of property from the street to the constitution. If the fundamental law of the state itself commands that key industries be publicly owned, that land belong to the state, and that insurance be a public monopoly, then the transfer does not await a revolutionary seizure, because there is nothing to seize: the constitution is already the supreme power, superior to legislature, executive, and judiciary alike, and officials who administer the scheme are executing the highest law rather than obeying a party’s decree. Entrenchment does the work that the dictatorship was supposed to do. A change of government cannot undo the scheme, because the scheme is not the government’s policy but the state’s constitution; a majority that wishes to restore private ownership must first amend the fundamental law under the special procedures the constitution prescribes, in full view of the electorate that benefits from the scheme. Add the franchise, which places legislative power, in the long run, in the hands of the propertyless many rather than the propertied few, and the mechanism is complete: persuasion creates the majority, the majority enacts the constitution, the constitution socialises property, and law holds the result against reaction. His slogan for the method, educate, agitate, organise, is the mechanism objection answered in three words: the road exists, and it runs through the minds of the majority rather than over the bodies of the minority.

He supported the construction with an argument about the comparative durability of legal and violent transfers that deserves to be taken seriously as political science. What is taken by force is held by force and lost when the force relaxes; what is enacted as fundamental law acquires the legitimacy that attaches to settled institutions, and each year of operation converts beneficiaries into defenders. He also pointed, in the lecture and elsewhere, to the internal evidence of the Buddhist case: the sangha had held its common property for centuries without a coercive apparatus, sustained by the conviction of its members and the authority of its rules, which suggested that common ownership resting on consent is not the contradiction the objection supposes. The state socialism of the memorandum generalises the sangha’s principle to the economy: common ownership as a rule of law, internalised by citizens who have voted for the order that embodies it. Whether the electorate of a poor and hierarchical country could be brought to enact such a constitution was, he acknowledged, a question of political labour; his parties, his newspapers, and his educational societies were the labour. The answer, in short, does not deny that power is required. It relocates power in the constitution and the franchise, and wagers that law, once made supreme, can do peacefully what the dictatorship promised to do by terror.

What is the strongest objection to his alternative?

That it lacks a mechanism against determined property. Persuasion and constitutional command, the Marxist respondents argued, leave ownership’s real defences untouched: wealth will capture the parties, the press, and the courts, and a socialist clause will be interpreted into harmlessness by the order it was meant to displace.

The mechanism question conceded

The discipline of this article requires the concession to be made in the open, as the brief for the debate directs: the question of mechanism is the weakest point in his alternative. The wager on law presumes that the constitution will be obeyed by the very interests it dispossesses, and his own career furnished grounds for doubt that he was too honest not to register. He had watched a reform bill he regarded as essential, the Hindu Code Bill, stall and die in the legislature of a free India whose constitution he had just drafted, and he resigned his ministry over the failure. If a parliament of the republic could not be brought to reform the marriage and inheritance law of the majority community, the critic is entitled to ask what would have happened to a clause socialising its land and industries. The amendment procedure he relied on cuts both ways: what one constituent moment entrenches, a later one, organised by frightened property with its superior resources, may dislodge, and the intervening administration of the scheme lies in hands the scheme does not choose. His construction has an answer to the question of legitimacy, since no transfer could be more legitimate than a constitutional one, and an answer to the question of durability in principle; to the question of enforcement against a propertied class determined to sabotage, capture, or outwait it, the answer is thinner, resting on the vigilance of an electorate whose education he knew to be the work of generations.

Later scholarship on both sides of the debate has tended to converge on this assessment even when it diverges on everything else. Writers sympathetic to his project treat the mechanism problem as the unfinished business of his thought, the point at which the programme awaits its engineers; writers in the Marxist tradition treat it as the proof that ends and means cannot be divided as cleanly as the lecture divides them, since a means inadequate to the end revises the end in practice. The convergence is itself a finding. The debate between Ambedkar and Marx, conducted at this level, is not a quarrel between a believer and an unbeliever but an argument about the physics of social change, in which both parties accept the destination and dispute the vehicle, and in which the dispute remains open because the experiment his memorandum proposed was never run. India socialised neither its land nor its key industries by constitutional command; it also never tested whether such a command, entrenched and majoritarian, could have survived its enemies. The historian’s verdict must therefore be the one the ledger suggests: the partial acceptance stands as the accurate description of his position, the mechanism objection stands as the accurate criticism of it, and the two together define the real debate that the slogans of both camps obscure.

Where does the mechanism objection leave his position?

Standing, but exposed at its joint. The division of Marx into accepted diagnosis and rejected method survives the criticism as an account of what he believed; what the criticism denies is that the substitute method, law and persuasion, could deliver the diagnosis’s cure against resistance.

Was he a socialist?

The query is among the most searched in the whole subject, and the texts permit a precise answer. He was a socialist in the sense that he held private ownership of the means of production to be a source of exploitation and suffering that must be ended by common ownership, and he drafted the most thorough scheme of state socialism ever placed before an Indian constitution-making body. He was not a socialist of the Marxist school, because he rejected that school’s philosophy, its theory of the state, and its method of transition, and he said so in terms. His own label, when he used one, was state socialism joined to political democracy: an economy socialised by constitutional law, governed by a democracy he insisted must become social and economic as well as political, and held together by a moral order he was prepared to call religion. The combination baffles taxonomies that assume socialism and constitutionalism exclude one another, which is a fact about the taxonomies rather than about him. He had read the European debate too well to suppose that common ownership required a single party, and he had read Indian society too closely to suppose that political democracy alone would dissolve its hierarchies.

Three clarifications protect the answer from its usual distortions. The first is that his socialism was not an election tactic of his later years but a position documented from the memorandum of 1947 backward into his economic writings, and consistent with the lecture’s endorsement of common ownership at the end of his life. The second is that his distance from the communist movement, documented in the charge section of this article, was a distance from an organisation and a method, not from the socialist objective, and conflating the two distances is the error that produces the portrait of him as an anti-socialist. The third is that his socialism was inseparable from the annihilation of caste: where the Marxist sequence ran from property to hierarchy, his ran in both directions at once, since a state socialism administered by a caste society would, in his analysis, be captured by the hierarchy it left standing. Socialise the factories and leave the village order untouched, and the twice-born will manage the factories. That sentence, which is his whole Indian argument in miniature, explains why his socialism had to be constitutional and moral as well as economic, and why it cannot be filed as a variant of anyone else’s doctrine. The related question of how caste and class compose a single system of oppression, which his position assumes rather than demonstrates, is pursued analytically in the companion article on the caste and class debate in India, and the history of the movement he opposed is told in the companion history of the communist movement in India.

Did he read Marx? The evidence of the engagement

A question that recurs in the forum literature asks whether Ambedkar had actually read Marx, or whether his strictures were aimed at a doctrine he knew only through its Indian salesmen. The question matters because the authority of his division of Marx depends on the division being made from the texts, and the evidence, happily, is not thin. Begin with the lecture itself. A man summarising a creed he has not read produces caricature; the summary in Buddha or Karl Marx is not caricature. It distinguishes the philosophical apparatus from the propositions that follow from it, identifies the redefinition of philosophy’s task as the creed’s starting point, and treats exploitation and private property as the load-bearing claims, which is how a careful reader of the Manifesto and of the first volume of Capital would arrange the matter. His criticisms are correspondingly aimed at real joints in the doctrine: the status of the economic interpretation, the logic of the transitional dictatorship, the prediction that religion would disappear. These are the points at which the Marxist tradition has argued with itself for a century, and a critic who finds them unaided has either read the texts or been very lucky in his informants. The lecture’s treatment of the Buddha shows the same habit of primary reading: its account of the sangha’s property rules is drawn from the monastic code he had studied for the book he was completing, and the comparison is constructed text against text throughout.

What is the strongest evidence that the engagement was firsthand?

The accuracy of the summary he attacks. His four propositions reproduce the creed’s real architecture, and his objections strike at the points the tradition itself treats as load-bearing: the economic interpretation, the transitional state, and the fate of religion under communism.

The circumstantial evidence supports the textual. He was, by the testimony of everyone who knew his habits, one of the most voracious readers in Indian public life: trained in two research universities, possessed of a private library that visitors described with awe, and accustomed to preparing for every controversy by mastering the opposing literature entire. His writings across the decades refer to the Marxist corpus and its Indian interpreters with the confidence of a man working from his own shelves, and his economic works engage the classical tradition, Smith, Mill, and their successors, in which Marx is a central figure, at the level of doctrine rather than slogan. Nor was the engagement a late discovery. The memorandum of 1947 presupposes a settled view of the socialist alternative; the polemics of the nineteen thirties and forties engage communist positions on labour and caste from close acquaintance with the movement’s publications; the lecture of 1956 crowns a familiarity of at least twenty years. The honest verdict is that he read Marx the way he read everything he intended to fight or to follow: completely, and with a pencil. Readers who wish to test the claim have a straightforward method available, which is to set his summary beside the sources and count the distortions. There are compressions, as in any summary of a system in four propositions, but the compressions are those of a advocate stating the case he is about to answer, and he is scrupulous, as the lecture’s structure shows, to declare his agreements before his disagreements. That order of proceeding is the signature of a reader, not of a pamphleteer working from extracts.

Liberty, equality, fraternity: the standard by which he judged both

Running beneath every section of this article is a triad that Ambedkar treated as the constitution of a free society, and making it explicit explains judgements that might otherwise seem miscellaneous. Liberty, equality, and fraternity appear in his work from the address of 1936 to the speeches of his final years as the three tests any proposed order must pass. A doctrine might promise equality and fail fraternity, as he held the dictatorship did, by organising society as a camp. It might promise liberty and fail equality, as he held the unreformed market order did, by leaving property’s power intact. The caste order failed all three at once, which is why he called it not merely unjust but anti-social: it denied liberty of occupation and movement, denied equality of status and opportunity, and destroyed fraternity by teaching men to regard one another as graded vessels of purity and pollution. The triad is drawn from the French revolutionary inheritance as he received and reworked it, and in his hands it is not a slogan but an instrument: three independent axes on which a social order can be plotted, so that a movement’s progress on one axis cannot be used to conceal its regression on another. Much confusion in the debate dissolves when the lecture is read as an application of the triad. Marxism scored high on equality of condition as an end and failed, in his judgement, on liberty during the transition and on fraternity as a method. The Buddha’s way scored on fraternity first, since it began by remaking men’s sentiments, and derived liberty and equality from the remaking.

Fraternity is the axis his interpreters most often slight, and it is the one he placed first in importance. In the address he argues that fraternity, the disposition to treat the interests of others as one’s own concern, is the root from which liberty and equality grow, because rights on paper are defended in practice only by a public that feels the violation of another’s rights as an injury to the common life. A society without fraternity will have liberty for the strong and equality before a law the strong administer. This ordering illuminates his strangest-seeming positions. His defence of religion, reconstructed as morality, is a defence of the institution that manufactures fraternity; his insistence on inter-caste marriage is a demand for the practice that embodies it; his suspicion of class politics conducted in indifference to caste is a charge that the movement lacks fraternity toward the very people it recruits. It also illuminates the warmth of his account of the sangha, an order whose entire rule was an engine of fraternity among men of every birth. When he said that the Buddha’s method was more permanent than Marx’s, he meant that an order built on fraternity renews itself in each generation’s sentiments, while an order built on coercion must renew its coercion. The triad is therefore not a decorative borrowing in his thought. It is the metric of the whole comparison, and the reader who holds it in view can reconstruct his verdict on any doctrine, including several he never lived to assess.

Why did he rank fraternity above the other two ideals?

Because he held it to be their root. Without the sentiment that makes another’s injury one’s own concern, liberty becomes the strong man’s charter and equality a form the strong administer; fraternity alone supplies the public feeling that defends rights in practice.

The lecture read closely: its argument in sequence

Because the lecture carries so much of the debate and is so often cited from fragments, a sequential reading is worthwhile, attending to what each movement of the argument does. It opens with a courtesy that is also a claim: the speaker, a convert of a few weeks, proposes to compare the founder of his new faith with the founder of the doctrine that ruled half the world, and to do so without heat. The creed of Marx is then summarised in the four propositions already examined, and the speaker declares which of them he accepts, pausing longest on the redefinition of philosophy, since it licenses the entire enterprise of reconstruction in which his own life was engaged. The turn to the Buddha follows, and its first move is chronological: the problem of suffering, and the insight that suffering is entangled with property and craving, were stated in India more than two thousand years before Marx, so that the comparison is between an original and a successor rather than between a sage and a scientist. The middle of the lecture conducts the institutional comparison through the sangha, whose property rules are described in detail drawn from the monastic code: the refusal of private possession, the common store, the government of the order by its assembled members. The point of the detail is argumentative. Each rule is an existence proof against a necessity claimed by the other side: common ownership without a state, equality without a party, discipline without a police.

The final movement addresses means and religion together, and it is where the lecture’s temperature rises. Force, he argues, can transfer property but cannot transfer allegiance; it can silence a doctrine but cannot refute it in the minds of the silenced; and a state constructed to wield such force will find reasons to wield it against the very class it liberated, since the machinery admits no distinction between enemies of the revolution and critics of its managers. Religion, correspondingly, is treated not as a survival to be outgrown but as a permanent need of social man, and the lecture closes by measuring the two teachers against the future each promised: one a harmony to be imposed when the state withered, the other a harmony to be practised as men were persuaded. The speaker awards the future to persuasion. Read whole, the lecture is less a verdict on Marx than a brief for a method, and its strictures on the doctrine are instrumental to that brief: materialism matters because it licenses force, force matters because it requires dictatorship, dictatorship matters because it cannot produce fraternity. A reader who extracts only the anti-Marxist sentences misses the brief; a reader who extracts only the agreements misses the verdict. The whole is the unit, which is why this article has insisted on the itemised form at every stage.

From the rupee to state socialism: the economist behind the programme

The memorandum’s scheme did not come from nowhere, and a short excursion into his formation as an economist explains its confidence. His doctoral work examined the finances of British India and the history of the rupee, and both studies are exercises in the same skill: tracing how an arrangement of money and taxation distributes power among classes and regions. The study of provincial finance showed how the imperial centre’s fiscal settlements structured what the provinces could do; the study of the rupee showed how a currency standard chosen in London bore on Indian prices, debtors, and peasants. He gave evidence before the royal commission on Indian currency in the nineteen twenties as a young expert, cross-examining the monetary orthodoxy of the empire in its own committee rooms. A man formed by such work did not think of property and money as natural facts. He thought of them as institutional choices, recorded in statutes and settlements, and therefore alterable by the same means. The road from the rupee problem to state socialism is shorter than it looks: in both, the question is who bears the cost of an arrangement, and whether the arrangement can be redesigned by law in favour of those who bear it.

His wartime years in the Viceroy’s council, holding the labour portfolio, added administration to theory. The portfolio placed him over labour policy in a wartime economy, and his tenure is associated in the record with the machinery of employment exchanges, with measures for labour welfare and social security in embryo, and with an insistence that the state treat labour as a constituency with claims rather than as a factor with a price. The experience matters for the debate because it was his one extended exercise of executive power before independence, and it confirmed the bent of his method: identify the institution, amend its rules, build the office that administers the amendment. When the Marxist respondents charged that his constitutional road had no mechanism, they were charging a man who had spent a career specifying mechanisms, down to the staffing of exchanges and the drafting of clauses. The charge survives, as the concession section of this article grants, at the level of enforcement against determined property; it does not survive as a portrait of a moralist innocent of administration. His alternative to the dictatorship was not a wish. It was the working method of a reforming minister extended to the scale of a constitution, and its weakness lies not in vagueness but in the wager that the method scales.

An indigenous radical tradition: Phule and the ground he stood on

Ambedkar’s confidence that India possessed resources of its own for the work of emancipation was not rhetorical. He stood in a line of anti-caste radicalism that began long before Marxism reached India, and he acknowledged the lineage explicitly by dedicating his study of the Shudras to the memory of Jotirao Phule, the nineteenth century Maharashtrian reformer who had founded schools for the lowest castes and for girls, organised the Satyashodhak Samaj against priestly authority, and recast the myths of the tradition as a history of conquest and subjugation. Phule matters to this article because he demonstrates that the critique of caste as a system, economic and religious together, was indigenous and prior: when Ambedkar charged that caste was guarded by scripture and reproduced in the family, he was extending an analysis his own tradition had begun before the Communist Manifesto had an Indian reader. The existence of that lineage shaped his posture toward Marxism in a way outsiders miss. He did not receive the doctrine as a thirsty man receives water. He received it as the heir of one radical tradition examining the claims of another, entitled to take what fitted his people’s case and to refuse what did not, and the entitlement explains the lecture’s tone of equality between the compared teachers. The Buddha, in that comparison, stands partly for the indigenous resource itself: proof that the subcontinent had generated its own doctrine of equality and its own discipline of transformation, which a modern movement could develop rather than import entire.

The lineage also clarifies the famous severity of the charge against the communist leadership. A movement heir to Phule’s tradition would have known, as a condition of its Indian birth, that caste was not a distraction from emancipation but its first object in India. The communists, by his account, treated that knowledge as provincial, subordinating it to a sequence derived from European experience in which the social question of hierarchy would be dissolved by the economic question of property. From within the indigenous line, the subordination looked less like science than like the old Brahminical postponement in modern dress: the untouchable was again asked to wait, this time for a revolution theorised elsewhere. His polemic’s heat comes from that resemblance. He was not refusing an ally’s help; he was refusing, for the third time in his career, an instruction to wait, from nationalists, reformers, and now revolutionaries, each of whom promised that his people’s question would be answered after the larger question was settled. The lecture’s calm and the polemic’s heat are the same conviction at different temperatures: that the annihilation of caste could not be sequenced after anything, because caste was the medium in which every other Indian question was suspended.

Accepting an end, refusing the means: the anatomy of a division

The partial acceptance is a more interesting intellectual structure than it first appears, and it rewards a moment of analysis in its own right, since examinations and essays on this subject founder precisely here. To divide a doctrine into an accepted diagnosis and a rejected method is to claim that the two are separable: that one can hold exploitation to be real, property to be its engine, and common ownership to be the end, while denying that the end requires, or is even compatible with, the philosophy and the instruments the doctrine prescribes. The Marxist tradition has generally denied the separability, on the ground that the method is deduced from the diagnosis: if the state is the executive committee of the possessing class, then persuasion of that class is a category error, and only a counter-power can dispossess it. Ambedkar’s division therefore contains an implicit counter-thesis about the state: that the modern democratic state is not reducible to a committee of owners, that its constitution can be captured by numbers and turned to purposes its original framers of property would abhor, and that law is a real force in history rather than a mask for force. The debate, at its deepest, is about what the state is. His evidence was the constitution he had just written, with its abolition of untouchability, its universal franchise granted at a stroke to a largely illiterate electorate, and its reservations of seats for the communities the hierarchy had excluded, instruments unintelligible on the committee theory of the state. His opponents’ evidence was the same constitution’s failure to touch land and industry, and the ease with which its social promises were administratively starved. Both exhibits are genuine, and the theoretical question they pose, whether law can be an autonomous lever of transformation or only registers transformations accomplished elsewhere, is the question the whole debate finally turns on.

What should a student separate between accepting an end and rejecting its proposed means?

Two judgements that slogans fuse. The end is judged by the values it embodies and the diagnosis that supports it; the means are judged by their side effects, their personnel, and their tendency to survive their purpose. A thinker may pass the first judgement and fail the second without contradiction.

For the student of political thought, the division also supplies a method for reading other encounters between traditions. Thinkers situated as Ambedkar was, heirs to an indigenous radicalism confronting a universal doctrine, characteristically divide what they receive: the pattern recurs wherever a doctrine travels, and the interesting question is never whether the receiver was for or against the doctrine entire but where the line of division falls and what principle draws it. In his case the line falls exactly where fraternity begins. Whatever in Marx could be affirmed without surrendering the moral agency of persons, he affirmed; whatever required persons to be treated as material for a process, he refused. The principle is stated nowhere in the lecture as a formula, which is what makes the division look miscellaneous in summary; but it predicts every row of the ledger, and a principle that predicts the evidence is the right candidate for the thought that produced it. His debate with Marx is therefore best taught not as a list of agreements and disagreements but as the working out of a single criterion across a system, and the series pattern this article instantiates, the partial acceptance, is the name for that working out.

The method in practice: tank, temple, text, and statute

His alternative method was not a theory in search of instances; the instances fill his biography, and reviewing them in sequence shows the method’s grammar. At Mahad in 1927 the issue was water: the municipality had declared the Chavadar tank open to all, and the untouchables, assembled in conference, marched to exercise the right in a body. The assertion was legal in form and revolutionary in content, since the right existed on paper and was denied in custom, and the method, mass presence under discipline invoking a public rule, was designed to expose exactly that gap. At the same conference a copy of the Manusmriti, the law book that codified the hierarchy’s theology, was publicly burned, an act of persuasion by symbol: the text that sanctified caste was treated as a political document and consumed as one. At Nashik from 1930 the issue was worship: the campaign to enter the Kalaram temple proceeded by satyagraha over years, accepting arrest and disappointment, until he concluded that temple entry into a religion that graded him was itself the wrong objective and turned his people’s energy toward exit and reconstruction. The sequence is the method learning: assert the right, test the institution, and where the institution’s theology is the offence, leave it and build another.

The constitutional phase extended the same grammar to the scale of the state. As labour member he built offices and amended rules; as law minister he attempted the most ambitious social legislation of the new republic, the Hindu Code Bill, which sought to remake marriage, inheritance, and adoption among Hindus on principles of equality, and whose strangulation in the legislature taught him the limit that the concession on mechanism, recorded in this article, describes. His educational work belongs in the same inventory: the People’s Education Society, founded in 1945, and the colleges it established were persuasion institutionalised, on his stated conviction that the oppressed needed the instruments of modern knowledge in their own hands. Even the conversion of 1956, the most dramatic act of his life, is legible in the grammar: a collective, announced, reasoned change of religious allegiance, executed by vow rather than by riot, and followed immediately by the composition of a scripture for the new community. Set beside the revolutionary method, the pattern is consistent enough to be called a doctrine of means: rights asserted in public under law, symbols contested by argument and ceremony, institutions built to outlast enthusiasm, and allegiance transferred by persuasion in mass. The Marxist objection, heard earlier, is that this grammar works only where the opponent permits it to work, and that property in arms will not be argued out of its arms. His counter-evidence was the abolition of untouchability in the fundamental law and the franchise in the hands of the formerly excluded. The argument between the two exhibits is the debate, and it is not settled by being summarised.

The same test applied to the national movement

The charge against the communists was one application of a test he applied to every movement that sought his people’s support, and seeing the test applied elsewhere protects the charge from being read as a factional grudge. To the national movement he put the same question he put to the Marxists: what happens to caste on the morning after your victory? His answer, argued across the Round Table period and codified in his 1945 indictment of Congress and Gandhi, was that independence under the existing social order would deliver the untouchables to the rule of the caste majority without the imperial referee, and that safeguards, separate representation and constitutional guarantees, were therefore the price of his people’s consent to the transfer of power. The Poona Pact of 1932, in which Gandhi’s fast compelled him to surrender separate electorates for reserved seats within joint electorates, was the formative defeat: he concluded that his people would henceforth be represented by men acceptable to the majority that oppressed them. Whatever the justice of that conclusion, its structural identity with the communist charge is plain. In both cases a universal movement, national or proletarian, asked the particular oppression to wait upon the universal victory, and promised that the victory would dissolve it. In both cases he answered that a hierarchy reproduced daily would survive any victory that did not aim at it directly, and that the promise to dissolve it later was the means by which the hierarchy’s beneficiaries within each movement protected their position now.

This consistency is the strongest internal evidence that the charge against the communists was analytical rather than personal. He had no more tenderness for the Congress leadership, whose caste composition resembled the communists’, and he applied the postponement test to Gandhi himself with a rigour that scandalised the national movement. The test also explains a feature of his thought that puzzles readers who meet him only in the Marx debate: his insistence on safeguards, reservations, and entrenched clauses, instruments that look legalistic beside the grandeur of revolution. They are the postponement test converted into institutions. A gain that can be postponed can be forgotten; a gain written into the fundamental law, staffed by the community’s own representatives and policed by its own organisations, has a chance of surviving the victory parades. The communist movement’s programme, as he read it, contained no such instruments for caste, because it did not regard caste as a structure requiring them. That omission, and not any quarrel over the distant end of a classless society, is what his charge finally names.

The state in his thought: trustee, instrument, and suspect

A thread that runs from the memorandum through the Assembly speeches to the lecture deserves gathering, because his theory of the state is the least summarised part of his position and the part on which the debate with Marx finally turns. He held three propositions about the state simultaneously, and the tension among them is deliberate. The state is an instrument: in a society of graded inequality, only an authority standing above the local powers of caste and property can enforce equality, which is why the untouchables’ emancipation required a strong constitutional centre rather than the romantic decentralisation favoured by the Gandhian wing of the national movement; the village, for him, was the hierarchy’s fortress, and the state its potential besieger. The state is a trustee: public ownership under his scheme is not the property of the government of the day but a trust administered under constitutional obligation, with the beneficiaries, the whole people, entitled to call the trustee to account through elections and courts. And the state is a suspect: power concentrated anywhere tends to perpetuate itself, which is why the trustee must be caged in fundamental law, divided among organs, and watched by a citizenry educated to fraternity. The dictatorship of the proletariat violated all three propositions at once in his reading: it made the state an instrument of one class, held the trust without a deed, and removed the cage on the promise that the animal inside was tame.

The Assembly warnings of his final period apply the same theory to the democracy he had helped to found. He cautioned the republic against the worship of heroes, telling the Assembly that devotion in politics is a road to degradation and eventual dictatorship; he warned that democracy in India was a top dressing on soil that was essentially undemocratic, meaning that the social habits of hierarchy would subvert the political forms unless the forms were used, deliberately and continuously, to remake the habits; and he insisted that the success of the constitution would depend less on its text than on the constitutional morality of those who worked it, a quality he defined as the disposition to restrain power by its own forms even when power could escape them. Read against the lecture, these warnings complete the alternative: the Buddha supplies the morality, the constitution supplies the forms, and each is the other’s safeguard. A Marxism reader may judge the construction naive, and the mechanism objection stands where this article placed it. But the construction is not the absence of a theory of the state. It is a rival theory, worked out against both the liberal night watchman and the proletarian dictatorship, and the debate is impoverished whenever it is summarised as socialism minus courage. What he feared was not change but the specific alchemy by which emancipatory power becomes a new hierarchy, an alchemy his people’s history had demonstrated under priest, landlord, and official, and against which he proposed the only prophylactics he believed existed: law above the rulers, morality within the citizens, and fraternity between them.

Why the answer was the Buddha and not liberalism

An examiner of the lecture sometimes asks why, having rejected so much of Marx, he reached for the Buddha rather than for the liberal tradition that also rejects dictatorship and violence and defends law. The question is fair and his work answers it. Liberalism, as he knew it from the inside of a caste society, defended the liberty of the person while leaving the person’s social degradation largely untouched: it offered the untouchable formal equality before a law whose substance was administered by the hierarchy, and it located the good society in the free play of individuals whom birth had equipped unequally for the game. His critique of that order runs through the address of 1936 and the Assembly speeches alike: liberty without fraternity is the liberty of the strong to despise in comfort. What he required from an alternative to Marx was therefore more than constitutionalism. He required a doctrine that attacked the sentiments of hierarchy directly, organised its adherents into a community that practised equality, and supplied a morality capable of governing conduct where law could not see. The Buddha, as he reconstructed him, supplied all three: a teaching that denied the authority of birth, an order that embodied the denial, and an ethics of compassion and friendship that reached into the kitchen and the marriage bed where caste actually lived. The choice was thus dictated by the diagnosis. If the enemy is a hierarchy of sentiments sanctified by religion, the counter-doctrine must be a religion reformed, not a philosophy of rights alone, and the lecture’s comparison is the public working of that requirement.

There is a second reason, of audience and allegiance, that should not be discounted. He spoke in the lecture to Asians and to Buddhists, and across his career to a people asked repeatedly to admire imported doctrines whose custodians despised them. To ground the alternative in the Buddha was to assert that the resources for India’s emancipation included India’s own deepest tradition of dissent, older than the caste order’s codification and untainted by it. The assertion had scholarly support in his reading of the texts and political meaning beyond scholarship: a movement of the despised that derives its charter from its own soil cannot be dismissed as a foreign contagion, the dismissal his opponents applied to Marxism and would gladly have applied to him. The comparison with Marx is therefore also a negotiation over inheritance. He takes from the European what the evidence commends, the diagnosis of exploitation and the end of common ownership, and he refuses the European custodianship of the future, locating the method and the morality in a teacher his own civilisation produced. Whether that negotiation persuades is for the reader; that it was deliberate is beyond serious question, and it is the final reason the partial acceptance should be read as a position rather than as indecision. He knew exactly what he was taking, what he was refusing, and in whose name he proposed to proceed.

The subsequent relationship: attempts to combine the traditions

Ambedkar died in 1956 with the debate formally unresolved: his charge unanswered in his lifetime at the level he had pitched it, his alternative untested, and the two traditions organised in separate and often hostile camps. From the nineteen seventies a series of attempts to combine the traditions appeared, and they are the empirical sequel to the lecture, since each attempt can be read as a judgement on where his division of Marx had been right and where it had been incomplete. The first and most vivid was literary and organisational at once. In 1972 a group of young Dalit writers and activists in Bombay founded the Dalit Panthers, taking their name and some of their manner from the Black Panthers across the Pacific, and their manifesto of the following year performed the combination in its definitions. A Dalit, for the manifesto, was not only the untouchable of the ritual order but the labourer, the landless peasant, the worker, the exploited of every description; the enemy was correspondingly double, caste and capital together, landlordism and Brahminism as one system with two faces. The Panthers read Marx and Ambedkar in the same room, quoted the lecture’s author on the annihilation of caste and the Manifesto on the abolition of property, and treated the combination not as a compromise between two orthodoxies but as the completion of each by the other. Their verse and their street politics carried the synthesis further in a few years than decades of party literature had done, and their internal quarrels, over whether class or caste held the primacy, rehearsed the debate’s unresolved question inside a single organisation.

Why did the Panthers reach for both traditions?

Because their constituency lived both oppressions at once. The landless labourer of a Maharashtra village was exploited as a worker and degraded as an untouchable by the same men, and a doctrine addressing only one of the two relations described only half of his morning.

The second attempt was theoretical and came from within the Marxist tradition itself. Sharad Patil, a scholar formed in the communist movement of Maharashtra, argued across a series of works that Indian Marxism had crippled itself by importing a sequence in which class preceded and explained caste, and he proposed instead a combined framework that he named for its three sources: Marx, Phule, and Ambedkar. In Patil’s construction, caste is not a survival draped over class relations but a constitutive structure of Indian society from its ancient formation, bound up with the control of labour and of women, so that an Indian materialism must begin where Phule and Ambedkar began, with the graded hierarchy as a primary fact, and read Marx through their corrections rather than despite them. The proposal matters because of who made it: not an Ambedkarite poaching on Marxist territory, but a Marxist concluding from inside that the movement’s caste theory required rebuilding around the very thinker his tradition had treated as a reformist diversion. Patil’s synthesis had its critics in both camps, which is the usual fate of syntheses, but its existence altered the terms of the debate: after it, the claim that the two traditions were simply incompatible had to be argued against a worked alternative rather than assumed.

What did Patil try to build from the two traditions?

A single materialism for Indian conditions. He held that caste was constitutive rather than residual, woven into the control of labour from the ancient formation onward, so Marx had to be read through Phule and Ambedkar’s corrections instead of around them.

The third attempt was scholarly and synthetic, and it is associated with Gail Omvedt, an American born researcher who settled in Maharashtra, married into the movement’s world, and spent a career writing the history of the anti-caste tradition as a democratic revolution in its own right. Her major study of the Dalit movement argued that the struggle against caste was not a preliminary to India’s democratic transformation but one of its central fronts, and that Marxism in India would remain a borrowed doctrine until it absorbed the anti-caste tradition’s analysis of a hierarchy that class categories alone could not map. Omvedt wrote explicitly as a Marxist working within the anti-caste inheritance, and her work performed the service the lecture’s debate most needed: it supplied the combined framework with footnotes, tracing the convergences through the non-Brahmin movement, the communist unions, and the Panthers, while registering the failures of each camp in the other’s terms. Her assessment of Ambedkar is correspondingly double. She treated him as the indispensable theorist of caste and as a democratic revolutionary whose suspicion of the party form was vindicated by the century, while pressing the mechanism question this article has conceded, the absence in his scheme of an account of how the propertied would be made to yield. The debate, in her hands, became a research programme rather than a polemic, which is perhaps the most that scholarship can do for a quarrel between the dead.

How did Omvedt restate the relationship?

As a completion rather than a compromise. The anti-caste movement was itself a democratic revolution, and Marxism in India would stay an imported doctrine until it absorbed that movement’s analysis of a hierarchy its class categories could not map alone.

Alongside these named attempts, quieter convergences accumulated in the record. The programmes of the communist parties themselves moved, across the decades after his death, toward explicit acknowledgement that caste oppression required direct struggle as part of the democratic programme, a formulation that would have been difficult to find in the manifestos he had prosecuted. Historians in the Marxist tradition produced work on caste and class, and on the ancient and medieval formations, that treated the hierarchy as a structure with its own weight, in the line of Kosambi rather than in the line of postponement. Ambedkarite organisations and publications, for their part, increasingly framed their demands in the language of labour, land, and public provision, and the figure of Ambedkar the economist, the author of the memorandum, was recovered from behind the figure of Ambedkar the constitution maker. None of this amounts to a merger, and the tensions the charge named did not evaporate: orthodox custodians on each side continued to police the boundary, and electoral competition for the same constituencies kept the quarrel warm in the districts. But the direction of the serious work is unmistakable across the half century after the lecture. Each tradition, at its best, moved toward the ground the other had made strongest: the Marxists toward the autonomy of caste, the Ambedkarites toward the economics of property, and the meeting ground they approached is recognisably the ground the partial acceptance had marked out, common ownership as the destination and the annihilation of caste as a co-equal condition of reaching it. The analytical development of that meeting ground belongs to the companion article on caste and class; its history belongs here, as the sequel that tests the lecture’s wager. The wager was that the two traditions needed one another. The sequel suggests he had read the future of both more clearly than either read itself.

Do Marxist and Ambedkarite organisations still keep their distance?

In part. Electoral rivalry and orthodox custodians in both camps maintained the boundary, and polemics continued; yet the serious work in each tradition converged on the ground the other held most strongly, and joint campaigns on land, labour, and atrocities recurred across the decades.

The scholarly assessment: biography, history, and the weighing of the debate

Between the polemics and the syntheses stands the scholarship that made the debate teachable, and a short account of it helps the reader place the article’s own procedure. His first substantial biographer, Dhananjay Keer, produced the life that fixed the narrative outline for a generation, working from papers and interviews while the events were within living memory; its value for this subject lies in the documentation of the reading, the campaigns, and the drafting years against which the texts must be read. Eleanor Zelliot, the American historian of the Mahar movement, spent a career on the social history beneath the debate, tracing how a community’s religious and political self understanding was remade between the nineteenth century and the conversion, and her work supplies the caution this article has tried to observe: that the doctrines were held by people for whom they were instruments of dignity before they were systems of thought. M. S. Gore’s sociological study of his ideology examined the internal coherence of the position, treating the acceptance of socialist ends and the rejection of communist means as a principled structure rather than an ambivalence, which is the judgement the ledger in this article formalises. The Marxist historiography, from Kosambi onward, supplied the ancient and medieval depth that Dange’s cruder genealogy lacked, and its general effect was to make the simple postponement position untenable inside the tradition itself: once caste had been shown at work in the organisation of production across two millennia, no serious materialism could promise its automatic dissolution by a change in the ownership of factories alone.

The weighing that emerges from this scholarship, across its camps, is remarkably consistent on three points. The first is that the engagement was real and firsthand, the point established in the evidence section of this article, and no longer seriously contested. The second is that the division of the doctrine was principled, drawn where his triad of liberty, equality, and fraternity required it to be drawn, and continuous across the twenty years between the address and the lecture. The third is that the debate’s open questions are the ones this article has flagged: the mechanism by which his constitutional road dispossesses, on which the Marxist objection retains its force, and the adequacy of class categories to a graded hierarchy, on which his objection retains its force. A debate in which each side’s strongest criticism of the other is still standing after seventy years is not a debate that has failed. It is a map of the two hardest problems in Indian social theory, drawn by the men who met them first, and the student’s task is to learn the map rather than to award the victory. The structural parallel with the debate over intersectionality and class, in which a further axis of oppression claims standing beside the economic, is drawn in the series companion and illuminates both: in each case the question is whether the added axis is a complication of the old map or a demand for a new one.

Misreadings that flatten the debate

Five misreadings recur in the coverage of this subject, and each is corrected by a portion of the record already set out. The first is that he rejected Marx wholesale. The lecture’s four acceptances refute it: a wholesale rejection does not endorse the enemy’s diagnosis, his account of property, and his definition of philosophy’s task in the same breath as its refusals. The second misreading is the mirror image, that he was himself a Marxist, a claim that survives only by suppressing the rejected half of the ledger, the philosophy, the violence, the dictatorship, and the religion, which together constitute most of what makes Marxism a system rather than a mood. The third misreading shrinks his objection to a grievance against the Indian communists, as though the doctrinal argument were an afterthought to a factional quarrel. The lecture refutes the shrinking in form as well as content: its addressee is the doctrine, its comparisons are with Marx and the Buddha, and its objections to materialism and dictatorship would stand entire if no Indian communist had ever existed. The fourth misreading treats the turn to Buddhism as a retreat from social analysis into private faith. The conversion’s own documents refute it: the lecture delivered in the same season is a work of comparative social theory, the scripture he composed for the converts is a social gospel organised around suffering and its causes, and the act itself was the largest collective assertion his movement ever made. A retreat does not usually take the form of leading one’s people across a frontier in public.

The fifth misreading is historical: that the two traditions have never combined, that the boundary drawn in his lifetime is the boundary of the subject. The record of the nineteen seventies onward refutes it, from the Panthers’ manifesto through Patil’s theoretical construction to the scholarly synthesis associated with Omvedt, and the refutation matters because the misreading does quiet work for both orthodoxies, licensing each to treat the other as a closed error rather than as an unfinished half. To these five a sixth might be added, of method rather than of fact: the habit of settling his position by anthology, quoting the sentence that serves the quoter’s camp. The antidote is the discipline this article has followed, to read the two governing texts entire, to keep the ledger’s rows separate, and to let his own order of exposition, agreements before disagreements, charge before replies, answer before concession, govern the presentation. A thinker who divided a doctrine with this much care deserves readers willing to keep the divisions intact.

Why is the wholesale rejection story the most durable of the errors?

Because it serves both camps. Marxists can file him as an opponent and Ambedkarites as a purist, and the filing spares each side the labour of answering the half of his position that accepts the other. The lecture’s declared agreements are the standing inconvenience.

Where to start with Ambedkar

The reading route through the debate follows his own order. Begin with the undelivered address, which is short, ferocious, and complete in itself: it gives the analysis of caste, the theory of religion’s guardianship, and the triad of ideals in their first full statement, and it explains why every later argument takes the shape it does. Read next the Kathmandu lecture, where the division of Marx is made in his own hand, and read it entire rather than in extracts, since its force lies in the sequence of agreements and refusals. The States and Minorities memorandum comes third, for the economic programme in drafted form, and it will correct any impression left by the lecture that his alternative was a sentiment: the clauses on land, industry, and insurance are as hard edged as anything in the period’s planning literature. The Buddha and His Dhamma, the scripture he composed for his converts and published after his death, comes last and longest; it is the positive doctrine, the reconstruction of religion as morality, against which the lecture’s talk of persuasion becomes concrete. The collected writings and speeches, issued in the official volumes by the government of Maharashtra across the decades after his death, gather the journalism, the evidence to commissions, and the legislative speeches that fill out the record, and the scholarly lives by Keer and the studies by Zelliot and Gore guide the reader who wants the man behind the positions.

For the student preparing on this subject, the instruments named in the series will serve better than any summary. The accepted and rejected ledger can be kept as a revision card in the study system at VaultBook’s Marxism study notes, where each row of the table becomes a prompt to be answered from the sources before it is checked against them. Candidates facing the civil service and eligibility examinations, for which this encounter is among the most examined topics in Indian political thought, will find the previous year questions gathered at ReportMedic’s UPSC previous year question explorer, and the questions repay sorting by the article’s own divisions: those that ask what he accepted, those that ask what he rejected, and those that ask the harder thing, why the line falls where it does. The field as a whole, the Indian career of the doctrine within which the debate sits, is surveyed in the series guide to Marxism in India complete guide, and the student who has placed the debate within that field is ready for the personas that close the article.

The earliest text: the mechanism of caste in the Columbia paper

Twenty years before the address, the mechanism that anchors his whole analysis was already worked out in a seminar paper at Columbia, presented in 1916 and published the following year in a journal of anthropology. The paper on castes in India asked how a caste, once formed, maintains itself in existence, and its answer was endogamy: the prohibition of marriage outside the group is the essence of caste, the one characteristic without which the rest is decoration, and the customs that surrounded it in Hindu practice, the burning of widows, enforced widowhood, and the marriage of girls before maturity, were explained as devices for disposing of the surplus man and the surplus woman whom strict endogamy periodically produced within a closed group. The argument is strikingly material for a paper of its date and milieu: it treats a religious institution as a machine with inputs, outputs, and waste problems, and it locates the machine’s working in the control of women’s sexuality and labour. It also displays the method that never left him: define the institution by its mechanism of reproduction, and you will know where to strike it. The address of 1936 strikes at endogamy with inter-caste marriage; the lecture of 1956 strikes at the philosophy that would leave endogamy’s guardians, the scriptures and the sentiments behind them, in place while rearranging property.

The paper matters for this article because it dates the independence of his analysis. When he later read the Marxist account of caste as superstructure, he was not encountering a rival speculation with a blank mind; he was measuring it against a mechanism he had specified as a student, from the ethnographic record, before the Russian revolution had occurred. The mechanism also explains the feature of his thought that most puzzles readers trained on class analysis: his conviction that caste could survive the socialisation of production indefinitely, because its reproduction runs through the household and the marriage bed rather than through the factory gate. A revolution that collectivised the mills of Bombay would leave untouched the arrangements by which a village in the Deccan decided whom its daughters might marry, and those arrangements, renewed each generation, would rebuild the graded order inside the new economy as reliably as they had inside the old. This is the concrete content of his theoretical disagreement with the superstructure doctrine, and it gives the disagreement an empirical address: the household. Later Marxist feminism and the social reproduction debates would pose cognate problems from within the tradition, about the labour and the institutions that reproduce labour power itself, and the series treats those debates in the feminism cluster; his paper of 1916 stands at the head of the Indian line of that questioning, which is one more reason the encounter with Marx should be read as a meeting of two developed analyses rather than as a prejudice meeting a science.

The texts on the other side: what in Marx he was answering

Fairness to the debate requires a sight of the positions he was answering in their own strongest textual form, since a rejection is only as serious as the version rejected. The dictatorship of the proletariat, in the texts he confronted, was not a licence for permanent one party rule but a transitional form modelled on the Paris Commune: a working class state, radically democratic in structure, armed for the suppression of the dispossessed minority’s resistance, and destined to lose its political character as classes disappeared. The phrase about religion, likewise, sits in a passage richer than its extraction suggests: religion in the 1844 introduction is the sigh of the oppressed creature and the heart of a heartless world before it is named the opium of the people, a formulation that grants religion the status of a protest against suffering as well as a narcotic against it. The economic interpretation, in its careful statements, claims that the mode of production conditions the general process of social, political, and intellectual life, a conditioning that its author’s collaborator later qualified with reciprocal action and relative independence. Ambedkar’s rejections were aimed at the doctrine as he found it operative in the movement and in the states that claimed it, rather than at every nuance the seminar room could restore, and a full record must note the aim. His defenders can reply, with justice, that a doctrine is responsible for its operative form: a transitional state that nowhere transitioned, and a conditioning that its parties read as determination, are facts about the doctrine’s career and not merely about its readers. His critics can reply, also with justice, that he sometimes argued against the career when he claimed to be arguing against the text. The lecture itself is more scrupulous than the polemics around it, since it summarises the creed in propositions a Marxist can recognise and answers them at that level; the movement charge then does the career’s work, in the proper place.

There is a parallel text that illuminates his performance by resemblance. In 1875 Marx himself had received the programme of a united German workers party and had subjected it to a marginal critique of legendary severity, accepting its socialist destination while rejecting its phrases, its theory of the state, and its silences about transition, and the Critique of the Gotha Programme, published after his death, is the tradition’s own model of the partial acceptance: loyalty to the end, ferocity toward the formula. Ambedkar’s division of Marx belongs recognisably to this genre of socialist criticism from within the family of emancipatory doctrines, however far he stood from the family’s organisation. He even shares with the Gotha critique a signature move: the insistence that distribution cannot be settled apart from the relations of production and the political forms that guard them, and the suspicion of programmes that promise the end while declining to specify the custody of power in the interval. To note the resemblance is not to conscript him into the tradition he declined to join. It is to identify the literary and intellectual form of his intervention, the programme critique, and to explain why the intervention retains its force for readers inside the tradition: it speaks the tradition’s own critical language back to it, with the caste question inserted where the silence had been.

Labour politics: unions, strikes, and the episodes of cooperation

The charge of caste-blind organising was made by a man with his own record in the labour movement, and the record complicates any picture of two camps that never met. He founded the Independent Labour Party in 1936 to carry a programme joining labour’s economic demands to the annihilation of caste in a single platform, and the party contested elections and sat in the Bombay legislature as a labour voice distinct from both Congress and the communists. His positions on strikes were characteristically double sided. He defended the right to strike as a civil liberty of workers and opposed legislation that would fetter it, joining the communists in 1938 in opposition to the Bombay government’s industrial disputes bill, which both movements read as a charter for suppressing stoppages; the one day protest strike that followed was an episode of practical cooperation between his organisation and theirs against a common legislative enemy. At the same time he criticised, in print and in the Assembly, stoppages called in the textile mills that he judged premature or conducted for the party’s ends rather than the workers’, and he charged the communist union leadership with spending the workers’ strength on gestures whose costs the workers bore. The two positions are one: strikes as a right are sacred, strikes as a tactic are answerable to the interests of those who strike, and a leadership that treats the mill hands as an infantry for its politics will hear from him in the same tone whether it sits in Delhi or in the party office.

These episodes are worth recording in an article on the debate because they calibrate the charge. The accusation of indifference to caste was not the reflex of a rival who had never shared a platform with the movement; it came from a labour politician who had cooperated with communists when the issue was a civil liberty, competed with them for the same constituencies when the issue was representation, and concluded from the inside of the same movement that its practice on caste was a dereliction. The cooperation also supplies the honest answer to one form of the question whether the traditions ever worked together: they did, episodically and against common opponents, within his lifetime and under his leadership, whenever the question at issue was one his test allowed to be common. Where the question was caste itself, the cooperation failed, because on that question the movements’ programmes diverged at the root; his demanded its annihilation as a first charge on any government, theirs subordinated it to a sequence he refused. The later combinations examined in the sequel section succeeded precisely where they repaired that divergence, by writing the annihilation into the common programme rather than appending it to the minutes. The history therefore teaches the same lesson as the lecture, in the grammar of events: the obstacle was never personal and never merely rhetorical, but programmatic, and it yielded only where programmes changed.

His judgement of the Soviet case

The lecture’s audience would have understood its references to the land where the Marxist method had been tried at full scale, and a modern reader should be given his position on that case distinctly, since the series discipline requires the theory and the state record to be kept in separate accounts. He did not deny the Soviet achievement in its own terms: the expropriation had been accomplished, a backward economy had been industrialised, and a form of equality of condition had been imposed across a vast society. His judgement concerned the price and the permanence. Equality maintained by a dictatorship, he argued, is held at the pleasure of the dictatorship, and a people that has been equalised by force has been taught obedience rather than fraternity, so that the equality itself is of the wrong kind: the equality of subjects, not of citizens. He drew the contrast with the sangha in terms of spiritual bookkeeping: the Buddhist order’s common life rested on the consent of men who could leave it, while the Soviet order’s rested on the impossibility of leaving, and the difference between consent and impossibility is the difference between the two methods in their entirety. On religion his judgement was predictive as well as moral: a state that set out to abolish religion would discover that it had abolished only its rivals, installing its own creed in the vacated sanctuary, complete with scriptures, prophets, and heresies. The prediction was made from observation of the movement’s habits rather than from theology, and it belongs to the record of the debate as one of its author’s most characteristic performances: the institutional analyst, watching a doctrine behave, and forecasting its behaviour from its structure rather than from its professions.

It should be recorded, in fairness to the separation of accounts, that his criticism of the Soviet case never became a defence of the order that opposed it. The same years produced his memorandum’s scheme for socialising the commanding heights of the Indian economy, a programme that would have made the Indian state the owner of more of the nation’s productive wealth than most socialist governments in Europe then administered. His quarrel with the Soviet model was a quarrel about custody and consent, not about the destination of common ownership, and readers who recruit him for the cold war’s other side recruit him falsely. He held, simultaneously and without embarrassment, that the Soviet Union had proved common ownership administrable and had proved dictatorship corrupting, and that an India wise enough to learn from the experiment would adopt the first lesson and refuse the second. That double judgement is the partial acceptance applied to the state record, and it closes the doctrinal part of this article where it began: with a thinker dividing what he was given, at the level of the institution and the argument, and declining to be filed under either of the banners that claimed him.

Judging movements by their social programme: the address on Ranade

His method of assessing a doctrine or a movement by its social content rather than its professions was set out programmatically in an address of 1943 on Ranade, Gandhi, and Jinnah, delivered at a Bombay college and published the same year. The address compares two nineteenth and twentieth century styles of Indian leadership: the social reformer who held that political advance without social reform was a house on sand, and the political leaders who subordinated every social question to the winning of power. His verdict, argued through the comparison, was that a movement is known by what it does to the social order it inherits: whether it attacks the hierarchy among its own followers, whether it makes the degraded castes and the subject sex its concern in the present or in the prologue to a future, and whether its professions survive contact with its own dinner table and marriage register. The address is the methodological charter for the communist charge a decade later. The movement was not being measured against an alien standard imported for the occasion; it was being measured against the standard he applied to every claimant to India’s future, Congress included and his own tradition not excepted, and the standard’s name is the refusal of postponement. A reader who knows the 1943 address will not mistake the charge for spleen. It is the output of an instrument, and the instrument can be checked against all of its other readings.

The address also discloses the intellectual virtue that organises his polemics and that examiners should learn to name: the insistence that ideas be audited in their institutional consequences. A philosophy, for him, is what its organisations do; a programme is what its clauses would do if enacted; a leader is what his movement does to the least of its followers. This is why the doctrinal and the movement criticisms, kept separate in this article as he kept them, nevertheless interlock in his practice: the doctrine is finally judged by the movement it produces when it is believed, and the movement is judged by the doctrine it invokes when it is questioned. Marxists who answer the charge by returning to the texts, and Ambedkarites who answer the doctrinal criticism by returning to the charge, are each using half of his own method against the other half. The whole method, texts audited by institutions and institutions by texts, is the standard this article has applied to both sides, and it is severe enough to have left its marks on every party it has touched, the author’s own alternative included, as the concession on mechanism records.

Land and the village: where the two programmes differed in rupees and acres

The debate becomes concrete at the agrarian question, and concreteness is a useful antidote to the exchange of adjectives. The Indian countryside he addressed was a structure of landlordism, moneylending, fragmented holdings, and hereditary village service, in which the untouchable typically stood as a landless labourer bound to the village establishment by customary obligation. The communist programme for this countryside, as developed in the movement’s agrarian documents, centred on the abolition of landlordism and the distribution of land to the tiller, confiscated without compensation, within a broader alliance of the peasantry against feudal and imperial power. His programme, in the memorandum, differed at the decisive joint. Distribution of land into small private holdings, in his analysis, would reproduce in economics the fragmentation he deplored in society: holdings too small to support a family or to use modern technique, a peasantry chained to dwarf farms and to the moneylender who financed them. He proposed instead that land be owned by the state, that cultivation be organised collectively by the village community with state capital and direction, and that the cultivator’s security rest on the collective’s charter rather than on a title deed to an uneconomic fragment. The proposal scandalised property and orthodoxy in equal measure, and it placed him, on this question, to the collectivist side of the movement that accused him of constitutional timidity.

The difference is instructive because it is not the difference the slogans predict. On land, the Marxist programme was the more individualist in its immediate form, vesting the peasant in his plot, while the constitutional lawyer proposed the more thorough socialisation, vesting the village in its land under the state. His reasoning joined his economics to his sociology. Economically, he had studied the uneconomic holding for decades and regarded its multiplication as a sentence of poverty. Sociologically, he distrusted the private plot as the material base on which the caste village would rebuild itself: give the hierarchy land titles, and it will sort them by birth within a generation. Collective cultivation under state ownership was designed to break the material link between caste and livelihood at the point where it was strongest, the village’s command over its labourers’ subsistence. The respondents’ rejoinder returned to the mechanism theme in agrarian dress: a state that owns the land of a caste society will administer it through the society’s existing officers, and the collective will be managed by the dominant caste in the state’s name, which is postponement’s mirror image, socialisation postponed by its own instrument. Both arguments are alive in the later record of Indian land policy, and the debate’s agrarian chapter deserves more attention than the anthologies give it, since it shows the two traditions reasoning about the same village with incompatible institutional imaginations: one trusting the peasant proprietor as the revolution’s beneficiary, the other trusting only the charter that stood above the village’s own powers.

Women, marriage, and the reproduction of the hierarchy

No account of his side of the debate is complete without the analysis of women, because his mechanism of caste runs through the control of marriage, and the control of marriage is the control of women. The 1916 paper had already drawn the conclusion in academic form: the customs that disfigured Hindu domestic life, the immolation of widows, the prohibition of widow remarriage, the marriage of infant girls, were not separate cruelties but components of a single system for managing the surplus persons a closed marriage market produces, and the system’s beneficiary was the caste order itself. His later essay on the rise and fall of the Hindu woman traced the same system’s theology, contrasting the relative standing of women in the early Buddhist centuries with their degradation under the codified law that followed, and assigning the degradation to the same priestly authority he fought elsewhere. The Hindu Code Bill of his ministerial years was the practical form of the analysis: a single legislative assault on the legal structure of the patriarchal family, covering marriage, divorce, adoption, inheritance, and the property of daughters, fought clause by clause through the Assembly and lost to the combined resistance of orthodoxy within and caution without. Its defeat and his resignation belong in this article for what they prove about his method’s limits, recorded already in the concession on mechanism, and for what they prove about his consistency: the man who told the communists that the household was a front line fought his own hardest parliamentary battle on exactly that line.

For the debate with Marx the significance is structural. The classical Marxist account, following Engels, treated the subjection of women as a consequence of private property, destined to be transformed as property was socialised and women entered social production. His analysis inverted the emphasis for the Indian case: the subjection of women was a mechanism of caste, which was itself a pillar of the whole order of property and status, so that an emancipation that left the marriage system and the family law untouched would leave the hierarchy’s reproductive machinery in working order. The two accounts are not so much contradictory as addressed to different joints of the same machine, and the later tradition of Marxist feminism, which the series examines in the feminism cluster, would spend decades working out their relation: how the household’s labour and the household’s rules reproduce the workforce and the hierarchy together. His priority in posing the Indian form of the problem is a matter of dates, since the mechanism paper precedes the works through which the question entered the Marxist mainstream in India, and it strengthens the article’s governing picture of the encounter: two analyses of one society, each with a working model of the part the other underweighted, and a debate that advances exactly to the extent that the models are combined.

Educate: the first word of the method

His slogan placed education first, before agitation and organisation, and the order was an argument. Persuasion as a method requires a persuaded class: men and women equipped to read the texts, audit the programmes, and staff the institutions the method builds. His own biography was the demonstration he offered his people, the untouchable boy at the schoolroom door become the drafter of the constitution, and he spent money, health, and organisational capital to generalise the demonstration: the hostels, the scholarships, the publishing ventures, and finally the People’s Education Society with its colleges in Bombay and Aurangabad, founded so that the community’s children would not depend on the charity of institutions that despised them. Education, in this scheme, is not uplift in the patronising sense. It is the manufacture of the citizenry that constitutional socialism requires, the slow creation of the majority whose conviction the mechanism depends on. The Marxist respondents’ taunt, that the method postpones emancipation to the pace of schooling, was one he had heard from every quarter and answered with the observation that the alternative pace, emancipation proclaimed by a committee, had its own delays and a heavier bill. The debate about tempo is genuine and unresolved, and the article records it as part of the mechanism question rather than as a separate quarrel. What is not in doubt is the coherence: a thinker whose method is persuasion must build schools as a revolutionary builds barracks, and the colleges bearing his name across Maharashtra are the material form of the lecture’s philosophy, as the memorandum is the material form of its economics.

After his death: the division inside his own house

The sequel inside his own movement supplies a final, ironic test of his division of Marx, because his followers proceeded to divide over it. The party he had founded in his last years, and the Republican Party of India that his lieutenants organised after his death to realise his design, carried his constitutional and parliamentary method into the republic’s politics, with his programme of social democracy and his suspicion of both Congress and the communists as their inheritance. But the generation that came of age in the slums and villages of the nineteen sixties and seventies found the parliamentary inheritance slow, and the Panthers’ turn to Marx, examined in the sequel section, was precisely a revolt within the house: the sons of his movement concluded that the father’s method lacked the mechanism his critics had always demanded, and reached for the rejected half of his ledger to supply it. The revolt was conducted in his name and against his text, which produced the curious literature of the period in which the lecture is cited by both the custodians and the rebels, the custodians quoting the rejections and the rebels the acceptances, exactly as the two national camps had done. The episode completes the article’s argument about the partial acceptance from an unexpected direction. A division so principled that it organised a great thinker’s relation to the strongest doctrine of his century also organised, after his death, the factions of his own movement, each seizing the half it needed. The historian’s task is not to reconcile the factions but to restore the whole from which the halves were torn, which is the task this article has attempted for the debate entire.

His theory of class, stated positively

The qualification he attached to the theory of class deserves a fuller statement than the ledger’s single row allows, because it amounts to a positive doctrine rather than a footnote. In the societies Marx analysed, class position is assigned by the market in labour and capital, and though birth heavily influences the assignment, the form of the system is mobility: the worker is formally free, sells his labour where he can, and may in principle rise or fall. In India, he argued, the assignment is made by birth into a caste, sanctified in advance, and enforced by the whole apparatus of custom; the result is that classes exist, but they exist as caste classes, orders in which the economic function and the ritual rank are fused in a single hereditary status. The fusion has consequences a class politics ignores at its peril. It fixes the personnel of exploitation across generations, so that the exploiting and exploited strata confront one another as communities with gods, customs, and marriage rules of their own, not merely as buyers and sellers of labour. It supplies the exploiters with a theology and the exploited with a counter theology of submission, karma and rebirth doing the work that ideology does in the European account, with the difference that the theology is not a reflection men can be argued out of by pamphlets about wages. And it divides the exploited among themselves by the same principle that unites them against the exploiter, so that the subjective condition of class politics, a common consciousness, must be manufactured against the grain of the workers’ own inherited identities. The series exposition of the Marxist theory of class guide sets out the doctrine in its classical form; his amendment is that in India the classical form describes the skeleton while caste supplies the flesh, and a politics addressed to the skeleton alone will be perpetually surprised by the body’s movements.

He pressed the amendment into a testable claim about consciousness. Where class and caste coincide, as they largely do at the bottom of the Indian order, the labourer’s consciousness of his condition arrives wearing caste’s clothing: he experiences his exploitation as his community’s degradation and his oppressor as another community before he experiences either as a relation of production. Movements that addressed the relation and ignored the clothing found their appeals translated, in the hearing of the audience, into the idiom they had refused to speak, and were then astonished when the audience followed the idiom’s own leaders. His own political practice was built on the claim. His parties organised the depressed classes as classes and as castes simultaneously, demanding land, wages, and education in one column and water, worship, and dignity in the other, on the theory that the two columns described one subjection. The respondents’ counter claim, heard in the replies section, that common struggle on the economic column dissolves the second column in practice, is the empirical heart of the whole dispute, and the record of the unions and the villages has been argued over ever since with exhibits on both sides. What cannot be argued is that the amendment was ignorant of the doctrine it amended. It is the work of a man proposing that the doctrine’s Indian edition requires a chapter its author never had occasion to write, which is a form of discipleship as well as of criticism, and one more reason the partial acceptance, rather than enmity or adherence, is the accurate name for the relation.

Democracy defined as change without bloodshed

His definition of democracy, given most sharply in the 1943 address and repeated in the Assembly years, deserves a section of its own because it is the hinge between his method and his ends. Democracy, in his formulation, is not primarily a machinery of voting; it is a form and a method of government by which revolutionary changes in the economic and social life of the people can be brought about without bloodshed. The definition is instrumental and audacious at once. It justifies the parliamentary form not by its moderation but by its revolutionary capacity: democracy is the technology by which a society transforms itself while keeping its members alive and its losers citizens. Measured by that definition, the dictatorship fails not because it is undemocratic in form, though he held that too, but because it purchases transformation at the price the definition exists to avoid, and then cannot guarantee the transformation once the price is paid. Measured by the same definition, a merely formal democracy, with votes above and hierarchy below, also fails, since it cannot deliver the revolutionary changes that are its warrant. His lifelong political project can be stated as the effort to make Indian democracy pass its own definition: to load the constitutional form with enough social content, franchise, safeguards, education, and eventually common ownership, that the form would actually perform the transformation ascribed to it. The effort explains his otherwise puzzling combination of reverence for the constitution and contempt for the society that received it. The constitution was the best instrument ever placed in his people’s hands; the society was the material the instrument had to cut; and his speeches of warning in the Assembly’s last days are the craftsman’s anxiety that the instrument would be blunted by the hands that held it, through hero worship, through the neglect of social democracy, and through the conversion of a method of change into a ritual of legitimation.

The definition also locates his difference from the liberal defence of democracy with final clarity. The liberal defends the form as the least bad arrangement for registering preferences; he defended it as the only arrangement yet devised for revolution without a charnel house, and he was correspondingly impatient with liberals who treated the form as an end and with revolutionaries who treated it as a fraud. Both, in his reading, had failed to ask what the form was for. The question gives his debate with Marx its last twist, because Marxism too claimed to want the revolutionary changes, and claimed indeed to be their science. The argument between them, at the very bottom, is an argument between two technologies of transformation claiming the same destination, and his case for his technology was empirical before it was moral: count what each had delivered, in the countries where each had been tried, for those at the order’s base. That is the comparison he invited in the lecture, and it remains the comparison by which the debate is most honestly continued.

The conversion as the alternative method enacted

The conversion of October 1956 is usually filed under religion, and it belongs equally in the history of political method, because it was the alternative to the Marxist route performed at full scale before the lecture that theorised it had been delivered. At Nagpur he and the vast gathering of his followers took refuge in the Buddha and his teaching and accepted a set of vows he had composed for the occasion, twenty two in number, which function as the constitution of the new community: they renounce the deities and rituals of the old order, reject the doctrine that any god was incarnate in the founder, forswear the rites for the dead that bound the living to the priesthood, and affirm the equality of all human beings and the determination to follow the path of compassion and wisdom. Read as a political act, the ceremony did in an afternoon what his theory said persuasion and law together could do: it transferred the allegiance of a people from a hierarchy that degraded them to an order that affirmed them, by vow rather than by compulsion, with the converts as the authors of their own transformation. No property changed hands at Nagpur and no office was seized; what changed was the sacred framework within which property and office would thereafter be judged by those who converted. The lecture at Kathmandu, delivered to the international custodians of the faith he had just joined, is the ceremony’s commentary: an explanation, to the old Buddhists, of what the new Buddhists intended, and of why their founder’s method and not Marx’s was the one they had enacted.

The vows repay attention as drafting, since drafting was his genius. They are not a creed of beliefs in the usual sense; they are mostly renunciations and resolutions, a bill of severance from the old order’s machinery itemised with a lawyer’s care: the gods named and dismissed, the priest disintermediated, the doctrine of birth based worth denied, the practices by which caste reproduced its theology, sacrifice, offering, and ancestral rite, individually foreclosed. Where a revolutionary party issues a programme for the state it hopes to capture, the vows issue a programme for the household and the heart, the two places where his analysis had always located caste’s real government. The conversion is therefore the strongest single exhibit for the defence of his method against the charge of mechanismlessness in the register of social reproduction: here is a mechanism, operating on allegiance and custom at mass scale, specified in twenty two clauses and executed without a casualty list. Its limits are the limits recorded elsewhere in this article: it changed no land register and nationalised no mill, and its author knew the difference, having spent the same years demanding that the state do by law what the vow did by faith. The completeness of his alternative lies in the combination, conversion for the sentiments, constitution for the structures, education for the capacities, and the debate with Marx is finally about whether that combination needed the additional element he refused. He died within weeks of the conversion and the lecture, leaving the combination as his answer, untested as a whole and unretracted in any part.

Parties, elections, and the record of the constitutional road

The constitutional road was not only theorised; it was travelled, and the travelling produced a record that both sides of the debate cite. His first party, the Independent Labour Party of 1936, won a striking bloc of seats in the Bombay legislature at the first provincial elections under the new franchise, on a programme joining labour demands to caste’s annihilation, and its legislators used the Assembly for the unglamorous work his method prescribed: questions, bills, amendments, and the exposure of administrative practice. The Scheduled Castes Federation, founded in 1942 as the political form narrowed to his community’s representation, carried the safeguards argument into the negotiations that preceded independence and into his memorandum to the Assembly. After independence the record darkens in ways his critics emphasise. He lost elections he expected to win, in Bombay in 1952 and in a by election two years later, defeated by the joint electorate system whose logic he had predicted at Poona: the representative of the depressed classes must be acceptable to the majority that is not depressed. The Republican Party, designed in his last years and founded by his followers after his death, inherited the design and the difficulty together. The critics’ reading of the record is that the constitutional road, travelled honestly for twenty years, delivered the constitution and the safeguards but not power, and that a method which cannot win power for its constituency will administer its emancipation as a permanent minority’s petition.

His defenders answer from within the same record. The constitution itself, with untouchability abolished, fundamental rights justiciable, and seats reserved in legislatures and services, was not a petition but a settlement, extracted at the one moment when his bargaining position was strongest and entrenched against the ordinary weather of politics; the electoral defeats occurred under rules he had opposed and predicted, and are evidence for his analysis of joint electorates rather than against his method; and the alternative road, travelled by the movement he criticised, delivered neither the annihilation of caste nor state power in the regions of his people’s greatest concentration, so the comparison of records does not award the palm where the critics assume. The exchange is recorded here without a victor, in the article’s discipline, because it restates the mechanism question in biographical form and because the honest reader should see that the debate’s evidence includes defeats on his side that his theory must digest, as it includes failures on the other side that theirs must digest. What the record does establish beyond dispute is the sincerity of the attempt: the constitutional road was not a debating position but a life’s work, and the lecture’s praise of persuasion was spoken by a man who had spent forty years and his health in the practice of it.

The newspapers: persuasion with a printing press

A method of persuasion needs instruments, and his first instrument was the press. In 1920, with support from the Maharaja of Kolhapur, he founded Mooknayak, the Leader of the Voiceless, a Marathi fortnightly that announced in its programme that the condition of the untouchables would be represented by the untouchables themselves. Bahishkrit Bharat followed in 1927, Janata in 1930, and Prabuddha Bharat, renamed in the year of the conversion, carried the line to the end of his life. The papers are the unglamorous infrastructure of everything this article has described: they reported the conferences and the satyagrahas, printed the speeches the metropolitan press ignored, argued the cases against Congress, reformers, and communists in a running polemical commentary, and educated a readership into the analysis its editor was elaborating in the books. For the debate with Marx they matter in two ways. First, they document the charge against the movement in its contemporary form: the editorials measured the communist unions and parties against the caste question as events unfolded, so that the indictment in the speeches rests on a paper trail of observation rather than on recollection. Second, they exhibit the alternative in operation. A newspaper is persuasion industrialised: argument repeated at intervals, addressed to readers free to disagree, building over years the public that a movement requires. The contrast he would have drawn is ready to hand. The revolutionary tradition also built great newspapers, and honoured the editor’s chair; the difference lies not in the instrument but in what the instrument was finally for, the capture of a state or the conversion of a society. His papers were aimed, with a consistency their files reward reading for, at the second of those objects, and they constitute the primary archive in which the debate was actually conducted for the audience that mattered most to him.

Broken men: his historical theory and the Marxist historians

His contributions to the theory of caste’s origin form the bridge between his sociology and the Marxist historiography he argued with, and they deserve record in any full account of the encounter. In The Untouchables, published in 1948, and in the unfinished manuscript on revolution and counter revolution in ancient India, he proposed that the untouchables originated as broken men: fragments of defeated tribes who, in the warfare of the early settlements, were broken from their kindred and attached themselves to the victorious villages as menials and watchmen, living outside the village wall in the quarters whose geography every Indian village still exhibits. The degradation of these already separate people into untouchables he assigned to a later ideological event: the triumph of Brahminism over Buddhism in what he called the counter revolution, after which beef eating, abandoned by the Buddhist influenced population at large and retained by the broken men who lived on carrion, was made the mark of a pollution so deep that touch itself defiled. Whatever the verdict of subsequent scholarship on the theory’s details, its form is the point for this article. It is a material and historical explanation of a religious status: war, settlement, diet, and doctrine interacting across centuries, with ideology doing real causal work at the decisive turn. A theory of that form is neither the idealism his Marxist critics sometimes attributed to him nor the economism he attributed to them. It is a research programme of the same family as Kosambi’s, differing in its assignments of weight, and the fact that the two men arrived independently at the project of explaining caste historically, from material beginnings, is the strongest evidence that the traditions were working one problem with two toolkits. The debate about ancient India between their successors, which belongs to the series treatment of the Indian Marxist historians, is the continuation of the argument by other means, and it began, as this article’s debate began, with each side holding part of a mechanism the other needed.

The Dewey inheritance: philosophy as reconstruction

The first of his acceptances, the redefinition of philosophy as reconstruction, had a source nearer to hand than Marx, and acknowledging it completes the map of his intellectual formation. At Columbia he had studied under John Dewey, and he acknowledged in later life the formative weight of that teaching: from Dewey he took the conception of democracy as a mode of associated living rather than a mere machinery of government, the tests of experience and consequences for ideas, and the habit of treating institutions as experiments answerable to the lives they organise. The address of 1936 quotes the pragmatist account of democracy with approval, and the triad of liberty, equality, and fraternity is in part Dewey’s associated living translated into the language of the French tradition. The inheritance matters for the Marx debate because it explains the otherwise surprising confidence with which he divided the doctrine. A reader formed by pragmatism is trained to ask of every proposition what difference it makes in practice and to keep the answer provisional on the evidence; such a reader will naturally take a system apart into its working claims, endorse those that pass the test of consequences, and discard the metaphysical wrapping that does no work. Dialectical materialism failed his examination partly as metaphysics that issued no instructions he could use, while the analysis of exploitation passed as an instruction he could act on the same week. The Buddha entered the same examination hall by the same door: the teaching was assessed as a technology of conduct and community, and it passed. Intellectual historians sometimes treat the Dewey connection as a softening influence, a liberal varnish on a radical. The record suggests the opposite: it was the source of the hardest edge in his method, the demand that every doctrine, Marx’s and the Buddha’s and his own, present its credentials in the currency of consequences for the people at the bottom of the order it proposed to change.

Safeguards: the social theory inside the clauses

The reserved seats, separate electorates, and constitutional guarantees that fill his political record are sometimes treated as a lawyer’s furniture, external to his social thought. They are better read as the social thought itself in institutional form, and a brief statement of their theory connects the debate’s halves. A safeguard, in his reasoning, answers a specific pathology of majoritarian society: where a minority is permanent, hereditary, and despised, the ordinary play of democracy will not merely neglect it but will select its representatives from among those the majority finds acceptable, converting representation into ventriloquism. The remedy is to alter the constituency of accountability, by separate electorates where possible and by reserved seats with a protected electorate where not, so that the representative answers to the represented. The theory generalises beyond its case. It holds that formal political equality is compatible with substantive political nullity wherever the social order pre sorts the plausible candidates, and that constitutions must therefore engineer against the social order’s tendencies rather than merely refrain from endorsing them. Applied to the Marx debate, the safeguard theory is the institutional corollary of the caste amendment to class theory: just as class consciousness arrives wearing caste’s clothing, democratic choice arrives wearing the hierarchy’s preferences, and a politics that will not design against the clothing will be governed by it. The communist respondents, with their suspicion of constitutional forms as bourgeois decoration, and the Congress gradualists, with their faith that the nation would outgrow its prejudices, stood together against this engineering, from opposite premises; his whole career is the argument that they were wrong together, and the safeguards in the constitution he drafted, reservations in legislatures, services, and education, are the argument’s surviving architecture, still load bearing and still contested on exactly the grounds he mapped.

The posthumous texts: an archive that changed the evidence

The debate’s evidentiary base widened considerably after his death, and the widening altered what could responsibly be claimed on every side. The scripture he composed for the converts appeared in print in 1957, a year after the conversion, giving the public the full text of the reconstruction that the lecture had assumed. A large historical manuscript on the ancient revolution and its Brahminical reversal, left unfinished at his death, reached print only decades later inside the official collected edition, and with it the fullest statement of his theory of how Buddhism fell and Brahminism triumphed, the theory on which the Broken Men account of untouchability’s origin depends. The collected writings and speeches themselves, issued volume by volume by the education department of the Maharashtra government from the late nineteen seventies onward, were the decisive event for scholarship: speeches once known from newspaper paraphrase could be checked against delivered texts, memoranda circulated in his lifetime to a few dozen officials became public documents, and the journalism of four newspapers entered the record in bulk. Positions in the debate that had rested on recollection migrated, volume by volume, onto the firmer ground of the page. The partial acceptance thesis itself is a child of that migration: it is difficult to sustain the portrait of a merely hostile critic once the lecture is read entire in a standard edition, and correspondingly difficult to sustain the portrait of a disciple once the rejected rows are read beside the accepted ones. Custodians of both camps have had to conduct their argument, ever since, in the presence of an archive that answers back.

The institutional afterlife: from polemic to syllabus

A further transformation overtook the debate as it passed from the custody of movements into the custody of institutions. Universities across India established chairs, centres, and courses bearing his name in the decades after his death; his portrait was installed in the central hall of the Parliament whose constitution he had drafted; and the two governing texts of this article entered the syllabi of political science and sociology as set works, to be read entire by students who arrive with no inheritance in the quarrel. The syllabus is an underrated arbiter in the history of ideas. Polemic rewards the extract, since a movement needs sentences that fit on banners; the seminar rewards the whole text, since a syllabus can assign forty pages as easily as four, and a generation taught from whole texts will find the anthology portraits of both camps increasingly difficult to recognise. Something of the same discipline reached the Marxist side of the shelf, where Kosambi’s histories and the later scholarship on caste and class became curricular in their turn, so that the replies too are now commonly met as arguments in books rather than as positions in a feud. The debate this article has reconstructed is therefore best understood as a living curriculum disguised as an old quarrel. Its questions, what caste is, what the state can be trusted to do, whether men are changed faster by conviction or by power, recur wherever the syllabus is taught, and they recur in his formulations because he stated them, in the address and the lecture, with a clarity that neither camp has improved upon. That is the afterlife the partial acceptance has earned: not a verdict filed away, but a standing examination that every Indian doctrine of emancipation is still required to sit.

The vocabulary the debate left behind

One measure of a debate’s weight is the vocabulary it deposits in the language of everyone who comes after, including its losers. Several of the phrases through which India now discusses hierarchy entered general circulation from his side of this argument. Annihilation of caste, the title he gave the address, fixed the objective at destruction rather than reform, and no subsequent programme in the field has been able to describe itself with a milder verb without explaining why. The figure of the ladder, a hierarchy ascending by reverence and descending by contempt, gave teachers a diagram for graded inequality that a shelf of monographs had failed to supply. His description of caste as a division among the workers themselves, rather than a mere allocation of jobs, passed into the working equipment of Indian sociology, where it is used by scholars who would not accept a single one of his remedies. Even the compound he coined for his remedy, state socialism fastened into a constitution, remains the reference point whenever an Indian programme proposes to settle an economic structure beyond the reach of ordinary legislation. The other side’s deposits are equally real: the description of caste as class relations wearing a sacred mask, the insistence that land and labour are the ground on which ritual rank finally stands, and Kosambi’s compressed formula for caste under primitive production all circulate in the same classrooms, often in the same lecture. A student meeting the subject for the first time is therefore already speaking the debate’s language before opening either author, which is why the article has laboured to restore the arguments beneath the phrases. Phrases that outlive their arguments become slogans; phrases reconnected to their arguments become instruments again, and the reconnection, for this pair of antagonists, is the whole of the work.

Two teachers, not two systems: the lecture’s chosen form

A peculiarity of the Kathmandu performance deserves a word of its own, because it governs how the whole piece should be read. He compares persons, a Buddha and a Marx, where a political scientist would compare systems, Buddhism and Marxism, and the choice is deliberate rather than quaint. Doctrines, in his handling, are inseparable from the communities that carry them and from the character of the carrying: a teaching is what its order does, its founder’s method is the prophecy of its followers’ conduct, and the surest way to judge a destination is to examine the road its first travellers actually walked. Hence the lecture’s persistent return to biography and institution, to the rules of an ancient mendicant order and the habits of a modern party, when a systematic treatise would have stayed with propositions. The form also performs a courtesy that the content requires. Setting teacher beside teacher places the European on the same bench as the Indian, neither as master nor as defendant, and allows agreements to be registered between equals before the disagreements are opened. Readers who find the pairing eccentric might consider what the alternative pairing, scripture against system, has produced in the century’s polemics: two fortresses exchanging summaries. His form produced something rarer, a judgement that partisans of the losing side could read without feeling caricatured, which is one reason the lecture, alone among the debate’s documents, is quoted with respect by scholars who reject its conclusion. Form, in this text, is not the clothing of the argument. It is the argument’s first premise, that how a doctrine is carried decides what it becomes, demonstrated by the carrying.

The pairing also fixes the scale at which the verdict is meant to operate. He is not asking which civilisation wins, nor which party a listener should join before the next election; he is asking which teacher a person bent on ending suffering should apprentice himself to, and the question is addressed to individuals who will have to act inside institutions they did not choose. That address explains the lecture’s survival among readers with no stake in either camp’s fortunes. Long after the organisations that first quarrelled over it have changed their names and programmes, the apprentice’s question remains on the table, in India and far beyond it, wherever a doctrine of liberation is proposed to people who must live under its methods as well as beneath its promises.

For students and exam candidates

Treat this debate as a set text problem, and the marks follow the texts. The two documents that decide every question are the undelivered address and the Kathmandu lecture, and the accepted and rejected ledger in this article should be memorised as a structure rather than as a list: agreements on the diagnosis and the end, refusals on the philosophy and the means. The guide to marxism for upsc guide places the topic inside the optional papers, and the worked answers collected in marxism in india exam answers show how the division is deployed under time pressure. In the examination room, state the partial acceptance in the first paragraph, name the respondents when the replies are used, and reserve a closing passage for the mechanism objection, which is where evaluation marks concentrate. Answers that declare him simply for or against Marx forfeit the topic’s central distinction and are graded accordingly by any examiner who has read the lecture.

For teachers and seminar leaders

The Kathmandu lecture teaches well because it is short, structured, and argumentative, and a seminar can be built by withholding its conclusion: distribute the four propositions he attributes to the creed, have the room predict which a leader of the untouchables would accept, and let the text correct the predictions. The address pairs with it as the earlier statement of the same mind, and the juxtaposition lets students watch a position hold its architecture across twenty years while its vocabulary changes. Set the replies alongside the charge rather than after it in the reading list, so that Dange and Kosambi are encountered as arguments and not as footnotes, and close the unit with the mechanism objection as an essay question in its own right, since it generalises to every programme that promises emancipation by instalments. The ledger table in this article can be projected as the seminar’s map, with each row assigned to a pair of students for verification against the sources, which teaches the habit this subject most needs: checking a summary against the text it summarises.

For researchers, historians, and journalists

The open questions are archival and comparative. His reading copies and marginalia, where they survive, would settle the depth of the encounter with particular Marxist texts in a way the published summaries cannot; the editorial files of his newspapers, incompletely mined, hold the contemporary form of the charge against the movement year by year; and the comparative question, how far the Indian debate anticipates the wider twentieth century argument about whether class categories can carry hierarchies of status, connects this material to the analytical treatment in the caste and class debate in India study and to the structural parallel examined in the series treatment of intersectionality and class. Journalists covering the periodic revivals of this quarrel in electoral seasons will serve their readers best by refusing the anthology method: the position is a structure, and any quotation that cannot be assigned to a row of the ledger, with its text and its year, should be treated as decoration rather than evidence.

The resignation: the method’s limit stated by its author

In September 1951 he resigned from the cabinet, and the statement he issued is among the most revealing of his documents for the purposes of this debate, because it is the method’s author auditing the method. He had entered the government, he explained, with two objects: to protect the interests of the scheduled castes and to carry the Hindu Code, the great codifying reform of Hindu personal law on which he had staked his ministerial reputation. The Code had been delayed, diluted, and finally shelved in the form he had drafted it, and he read the shelving as a verdict on the instrument he had trusted: a parliament representing a society could be brought to declare rights in the abstract and would still refuse, when the clauses reached the household, to legislate against the customs its members lived by. The statement is not a recantation of the constitutional road; he continued to the end of his life to regard the constitution as the community’s greatest possession. It is a calibration of it. Law, he had learned in office, moves at the speed of the social conscience behind it, and where the conscience refused, the clause waited. The lesson fed directly into the final phase of his work: if the state alone could not carry the transformation, the sentiments of society had to be carried too, which is the argument for the conversion conducted five years later and for the theory of religion that the lecture defends. The Marxist respondents read the resignation differently, as the mechanism objection’s documentary proof: the state had declined to dispossess even in the small change of inheritance law. Both readings are recorded in the debate’s literature, and the document sustains each of them in part, which is why it belongs in any honest account of where his alternative was strong and where it bled.

Linguistic states and the engineering of majorities

A late work of apparently narrow scope, his analysis of linguistic states written in 1955, shows the institutional imagination at work on a fresh problem and confirms that the safeguard theory was a general doctrine rather than a caste expedient. Called to consider the reorganisation of India’s provinces by language, he supported the principle and then surrounded it with conditions drawn from his lifelong study of majorities: a linguistic province creates a regional majority with a language, a literature, and a bureaucracy of its own, and minorities inside it, linguistic, religious, and caste, require constitutional protection against the new majority’s natural confidence. He proposed safeguards accordingly, including provisions for the official recognition of minority languages and the siting of capitals with an eye to the balance of communities. The parallel with the Marx debate is exact. In both cases he refuses the sequence in which a large emancipation, national reorganisation or social revolution, is trusted to dissolve the smaller dominations inside it by its own momentum. Dominations, in his sociology, do not dissolve; they relocate, and a constitution that does not design against the relocation has merely chosen the next address of the old hierarchy. The respondents in the Marxist tradition prized the large emancipation and treated the smaller dominations as its unfinished business; he treated the unfinished business as the main text. That difference of emphasis, repeated across land, language, religion, and caste, is the most portable lesson the debate offers to students of other countries’ revolutions, and it is why his name recurs wherever a movement is accused of asking a minority to wait.

The long deliberation: why Buddhism

The choice enacted in 1956 had been deliberated for twenty years, and the deliberation illuminates the tests his theory of religion imposed. After the declaration of 1935 that he would not die a Hindu, he examined the available faiths as a constitutional lawyer examines constitutions: Sikhism was explored seriously in the years that followed, with emissaries and negotiations that the record preserves; Christianity and Islam were weighed; and each was tested against the criteria the lecture later states openly. The adopted religion had to affirm human equality without remainder; it had to be compatible with reason, making no demand that its adherents believe against the evidence of their senses; it had to be capable of binding its adherents to fraternity in conduct; and, in the Indian setting, it had to avoid delivering his people into a new communal subordination or marking them as the clients of a foreign power. Buddhism, as he reconstructed it, passed the tests its rivals failed, with the added authority of an Indian origin and a founder who had himself broken with the priestly order of his day. The reconstruction was radical, and he knew it: the scripture he composed strips the teaching of the supernatural machinery that popular Buddhism had accumulated, recasts karma and rebirth in moral and social terms, and presents the founder as a teacher of society rather than a worker of wonders. Orthodox custodians of the tradition have disputed the result ever since, and the dispute belongs to religious studies; for the debate with Marx, what matters is the procedure. Faced with the need his theory asserted, for a religion that could carry a morality of equality, he did not wait for one to evolve. He selected, edited, and promulgated one, with the same drafting confidence he brought to a constitution, and the act is the boldest exercise of the persuasion method in his career: an attempt to change what millions of people held sacred by giving them a text and a vow and an argument, all at once.

The Pakistan study: the method applied to another hierarchy of belonging

His study of the Pakistan question, issued in 1940 and enlarged in 1945, is usually shelved with the literature of partition, but its method is the method of this article’s debate and a paragraph belongs to it. Confronted with the demand for a separate Muslim homeland, he declined both the reflexes of his contemporaries, denunciation and dismissal, and subjected the demand to an audit of the same kind he applied to programmes: what social facts sustained it, what safeguards were proposed within a united alternative, and whether those safeguards could command the confidence of the minority they were meant to protect. His conclusion, that the demand expressed a real failure of the majority to make safeguards credible, scandalised the national movement precisely because it applied his test impartially: a minority that is asked to trust the future goodwill of a permanent majority is being asked to accept the postponement structure in constitutional dress, and he had spent his life refusing that structure on behalf of a smaller and weaker minority than the one then in question. The study matters here for two reasons. It demonstrates that the postponement test was not manufactured for the communist case but was the constant instrument of his political judgement from 1932 onward, applied to Hindu, Muslim, and Marxist majorities with the same calibration. And it shows the caste argument’s root system: beneath the analysis of caste lies a general theory of how permanent minorities fare inside democratic forms, and the theory, not the community’s interest alone, is what he carried into the encounter with Marx. A doctrine that proposed to abolish minorities’ fears by abolishing classes had to persuade a man whose science was the fear of permanent minorities; the lecture’s scepticism about the withering state is that science speaking.

The Assembly speeches: the social democracy still owed

His two great addresses to the Constituent Assembly bookend the republic’s founding and state his alternative in its most public form, and a reader of the debate should have them in view. Introducing the draft constitution in November 1948, he defended its design against critics from every side and reserved his sharpest scorn for the romantic cult of the Indian village, the supposed republic of the nationalist imagination: the village, he told the Assembly, was a sink of localism and a den of ignorance, narrow mindedness, and communalism, and the constitution did well to take the individual, not the village, as its unit. The sentence is his sociology in an epigram. The romantic village was the caste order’s home address, and a constitution built upon it would have constitutionalised the hierarchy; a constitution built upon the citizen could, in time, dissolve it. A year later, moving the final adoption, he delivered the warnings this article has already used, about hero worship as the road to dictatorship and about the contradiction between political equality and social and economic inequality, which he said could put the political democracy itself in peril if it were not removed. He also addressed method, in terms that reach back to the lecture and forward to every Indian movement since: the methods of civil disobedience and non cooperation, legitimate against a foreign despotism, become in a constitutional order what he called the grammar of anarchy, since they short circuit the machinery by which a free people is supposed to change its laws. The argument completes his doctrine of means from the side of the citizen. Not only must the state refuse dictatorship; the movement must refuse the shortcut, accepting the discipline of the forms it claims to defend, because the forms are not the enemy of the transformation but its only durable vehicle. Marxist critics read that passage then and since as the constitutionalist’s surrender dressed as principle. His supporters read it as the same insight as the lecture’s, applied at home: that means are not laundered by ends, in the street any more than in the state.

Social endosmosis: the picture of the free society

The positive image behind the address of 1936 is a piece of sociology with a borrowed biological name, and naming it gives the debate its final analytical term. An ideal society, he wrote, would be mobile and full of channels for conveying change, with many interests consciously shared and communicated, and points of contact with other modes of association varied and numerous: in the word he adopted, social endosmosis, the free percolation of persons, ideas, and sympathies across the whole body of society. Caste is the precise negation of the image. It organises society into sealed vessels, each with its own occupations, gods, and marriage pool, between which no healthy circulation occurs; what passes between them is not sympathy but command downward and deference upward. Every element of his position can be restated in the image’s terms. Exploitation persists because the vessels are sealed: the suffering of one compartment is not felt as a common injury, which is the failure of fraternity. Class politics fails because the compartments cut across the classes the politics assumes. Religion must be reconstructed because the vessels’ seals are sacred. Persuasion is the required method because seals of belief are broken by teaching and example, not by confiscation. And the sangha commended itself because it was, in his reading, the one ancient Indian institution built on endosmosis: entry open, property common, rank by seniority of commitment rather than by birth. The image also supplies the fairest test of the Marxist counter proposal on its own ground. A revolution that smashed the seals by force while sealing its own party against society would have changed the arrangement of the vessels rather than dissolved them, and the history he surveyed in the lecture seemed to him to show exactly that substitution. Endosmosis cannot be decreed, which is his point; it can only be grown, by law that forbids the seals, education that ignores them, marriage that crosses them, and a morality that counts them sins. The slowness of the growing is the standing objection, and it stands; but the image at least specifies what the slow work is for, which is more than the metaphors of seizure ever did.

The village services: untouchability as an economic system

His analysis of caste was never only doctrinal, and its most concrete stratum is his account of the village service system under which his own community laboured. In the Maharashtra of his birth, the Mahars held a hereditary village office, the watan, which bound them to a schedule of duties for the village and its officials: carrying messages, guarding the village, removing dead cattle, serving touring officers, and performing the miscellaneous labour of the settlement, in return for a small grant of land and customary dues in kind from the cultivators. The arrangement wore the dress of an office and the reality of a bond. The duties were compulsory, the remuneration customary and mean, the tenure hereditary, and the obligation enforceable by the whole weight of the village order; a Mahar who declined the work declined his caste’s assigned place and could be made to feel the declension in every relation of his life. He campaigned against the system from the nineteen twenties, introduced legislation in the Bombay Legislative Council to commute and reform the watans, and returned to the charge across his career, because the watan was his standing proof that caste in India was not a prejudice draped over an economy but an economy itself: a mechanism for extracting labour without a labour market, secured by ritual status and land law together.

The watan matters to the debate with Marx for the precision it gives his amendment to class theory. Here is exploitation in its purest institutional form, surplus labour extracted by hereditary right, and yet it cannot be described in the vocabulary of capital and wage labour without remainder. The extractor is not an employer but a village order; the sanction is not the labour contract but the scripture and the custom; the labourer’s unfreedom is not the economic compulsion of the propertyless but the ritual compulsion of the polluted. A programme for abolishing the watan by socialising the means of production would have to specify which means, in a village of hereditary offices and customary dues, it proposed to socialise, and the specification is the whole difficulty. His own programme specified it without flinching: state ownership of land and the reorganisation of the village’s labour under public law, dissolving the hereditary offices into salaried public functions. The Marxist agrarian programme, abolishing landlordism and distributing land, addressed the village’s summit, the landlord, more directly than its base, the service order, and his criticism was that the base would survive the summit’s fall and reconstitute a hierarchy on the distributed plots. The later history of land reform in the region, in which ceiling and tenancy legislation coexisted for decades with caste’s command of village labour, is the archive in which that criticism and its rivals are still being tried, and the student of the debate should know that it began not in a text about Marx but in a bill about watchmen’s dues.

The party question: representation against vanguard

His organisational practice embodied a theory of the party opposed, point by point, to the vanguard conception he rejected in the doctrine. A party, in his construction, is the organised form of a defined constituency’s interests, built to represent that constituency in the arenas where its fate is decided, and accountable to it through the ordinary disciplines of membership, programme, and election. Its leader is its servant and its draftsman, not its prophet; its programme is a public instrument to be judged by its clauses; and its success is measured by the safeguards and statutes it extracts, not by the purity of its doctrine. The parties he founded observe the theory in their documents: the Independent Labour Party with a programme of labour and caste demands itemised for legislation, the Scheduled Castes Federation organised around representation and safeguards, the Republican design left as his final organisational testament. He was unsparing about the contrary model, in which a party constitutes itself the trustee of a class’s historical mission, interprets the mission through its central committee, and treats the class’s actual organisations as transmission belts for decisions taken elsewhere. Such a party, he held, reproduces inside the movement the structure of the society it opposes: a priesthood of interpretation above, a laity of obedience below. The charge against the Indian movement’s caste composition draws part of its force from this theory, since a vanguard recruited from the twice born and armed with a doctrine that postponed caste was doubly insulated against correction by the constituency it claimed to serve. His alternative insulation was deliberate exposure: leaders answerable to the led, programmes printed and priced, and a press in the community’s own hands to conduct the audit.

It is only fair to record the tension inside his own practice, since the debate’s discipline of concessions applies to its subject as well. The movement he led was intensely personal in its authority, and his followers’ devotion to him was of a kind he himself warned assemblies against when he spoke of devotion in politics as a road to degradation. He knew the contradiction and spoke to it: the safeguards he demanded in the constitution were demanded, in part, against the day when his own community’s gratitude would make it careless of its representatives’ quality. Whether the organisations he left behind solved the problem of succession and internal democracy is a question their subsequent history answers imperfectly, and his critics in both camps have used the imperfection. For the debate with Marx the relevant point is narrower and stands: his objection to the vanguard was not an objection to organisation, of which he built more than most of his critics combined, but to a specific theory of the party’s relation to truth, in which the party owns the doctrine and the class must be educated to deserve it. In his counter theory the constituency owns the programme and the party must be educated, election by election and clause by clause, to deserve the constituency. The two theories generate different movements, different newspapers, different failures, and, when they succeed, different republics.

Bombay and the Deccan: the ground on which the charge was formed

The debate’s abstractions had a street address, and supplying it disciplines both sides’ claims. The Bombay of the nineteen twenties and thirties contained, within a few square miles, the largest concentration of textile mills in the country, a communist movement of genuine strength among the mill hands, and the headquarters of Ambedkar’s organisations and newspapers. The great textile strikes of 1928 and 1929, led by the communist Girni Kamgar Union, were the movement’s finest hours in the city: tens of thousands of workers held out for months against the mill owners, and the union’s organisers, Dange among them, built a reputation for courage and discipline that the movement’s later literature rightly honours. It was against this background, of a movement he had watched at close quarters in its strength, that his criticism of its caste practice was formed, and the criticism is precise in proportion to the proximity. The mills drew their labour from the countryside in caste patterned streams; departments within the mills were sorted by community, with the better paid and cleaner work of the weaving sheds effectively closed to untouchable workers, who clustered in the lower paid spinning and waste work; and the union, organising the mill as it found it, fought the owners over wages and rationalisation while leaving the departmental colour bar of caste substantially unchallenged, since to challenge it would have divided the workforce on the eve of every strike. He named that omission, in the Assembly and in print, as the movement’s Indian signature: solidarity against the owner, silence before the foreman’s custom.

The respondents’ counter record, to be weighed in the same streets, is not empty. The unions did unite workers of every caste in common action on a scale no other organisation of the period achieved, and the experience of the picket line and the strike kitchen was itself a school of equality that participants remembered in those terms. Communist organisers lived in the chawls, shared the workers’ conditions, and in the famine of welfare provision in the mill districts built the beginnings of clinics and reading rooms. His charge was never that the movement did nothing; it was that what it did stopped at the mill gate of caste, and that the stopping was principled, derived from a theory that assigned caste to the superstructure and a personnel formed by the hierarchy the theory neglected. The Deccan background completes the picture. In the countryside behind the city, the non Brahmin movement and the patronage of the Maharaja of Kolhapur, who had funded his first newspaper and his students’ hostels, had built an anti-caste public sphere a generation before the communists arrived in strength, and his politics grew in that sphere’s soil: its premise was that caste had to be fought as itself, by organisation, education, and law, in the village and the school as much as in the factory. The two movements thus confronted one another as products of two different Indian experiences of oppression, the industrial and the agrarian ritual, each generalising its own experience into a theory of the whole. Their debate, which this article has followed through texts and programmes, is at bottom that confrontation of experiences, and it explains the debate’s durability better than any doctrinal summary: the mill and the village are both still there in the Indian social formation, and any theory that begins from only one of them will keep meeting the other’s objection.

The wider field: Congress socialists and the other claimants

The debate with Marx was conducted in a crowded field, and the crowd explains some of its pressure. Indian socialism in his lifetime was not a single doctrine but a bazaar of claimants: the communist parties with their international discipline, the Congress Socialist Party founded in 1934 within the national movement by a younger generation impatient with its leadership’s caution, independent figures such as M. N. Roy who passed from founding communism in exile to a radical humanism of reason and decentralised democracy, and the Gandhian constructive programme with its own economics of the village and the spinning wheel. His judgement of the Congress socialists applied the postponement test with the same calibration as his judgement of the communists, and found the same fault in a milder form: a socialism nested inside a national movement whose unity depended on the caste order’s tranquillity would postpone the annihilation of caste until after independence, and then, in government, discover reasons of state for postponing it further. The reservation of his sharpest polemic for the communists, given this wider field of disappointment, follows from the stakes rather than from proximity alone. The communists were the claimants whose doctrine he had largely accepted, whose organisation in the mills was strongest, and whose failure on caste therefore carried the largest consequence: if the most serious socialist movement in India could not be brought to fight the hierarchy, the hierarchy would survive socialism’s victory as it had survived every previous change of management. His relations with Roy’s later humanism, by contrast, show the tests working in the affirmative direction: a doctrine that placed reason, morality, and the individual at its centre, and renounced dictatorship, could meet his criteria in large part, and the record shows a corresponding respect between the two men in their later years, across the distance of their organisations. The field’s map, so drawn, is the partial acceptance generalised: he sorted every claimant by the same criteria, the diagnosis it offered, the means it proposed, and the place of caste in its programme, and the communists’ tragedy in his account was to pass the first test more completely than any rival and fail the third more consequentially.

Frequently Asked Questions

Q: Who was B.R. Ambedkar?

Bhimrao Ramji Ambedkar (1891 to 1956) was born at Mhow into a Mahar family, a community the Hindu order placed beyond the pale of touch. Scholarships and ferocious application carried him to doctorates at Columbia and in London and to the English Bar. He returned to lead the untouchables’ movement: conferences, newspapers, satyagrahas for water and temple entry, parties built for representation, a wartime labour portfolio, and then the law ministry of free India, from which he steered the drafting of the republic’s constitution. In the last weeks of his life he led his followers into Buddhism, an exit he had announced two decades earlier.

Q: Did Ambedkar read Marx?

The lecture of his final year settles the question for any fair reader. Its summary of the creed is compressed but accurate, arranged around the claims that genuinely carry weight in the system, and its criticisms land where the tradition’s own internal debates have always been fiercest: how far economic life explains the rest, whether a transitional dictatorship can be trusted to end, and what becomes of religion when exploitation ends. A polemicist working from extracts does not find those joints. Add the setting of his life, two research universities, a famous private library, decades of printed controversy with the movement’s journals, and the conclusion is unavoidable: the reading was direct and deep.

Q: What did Ambedkar accept from Marx?

He said so himself, in a numbered spirit if not in numbers. Philosophy, he agreed, is for remaking the world, not for glossing it. Classes exist and their interests collide. The collision is not a misunderstanding: it is exploitation, the systematic taking of labour’s product by the owners of the conditions of labour. And the root of the taking is private ownership of productive wealth, which arms one class with power and sentences the rest to suffering, so that a decent society must abolish the arrangement rather than decorate it. His 1947 memorandum shows the acceptance was no courtesy: it wrote public ownership of land, of the commanding industries, and of insurance into a draft constitution.

Q: What did Ambedkar reject in Marxism?

Begin with the metaphysics: he would not grant that matter and its motions exhaust reality, nor that the creed built on that grant behaved, among its believers, like anything but a faith. Continue with the instruments. Bloodshed as a method he refused, on the ground that a society founded by terror must keep terror on the payroll. The transitional dictatorship he refused more sharply still, asking who would wind it up and receiving, he thought, a prophecy in place of an answer. And he refused the promised disappearance of religion, holding that men will always need a sacred moral order and that the task is to rebuild it on equality, not to vacate the shrine for worse tenants.

Q: What was Ambedkar’s criticism of the communists?

It was an indictment in three counts, and he delivered it as a prosecutor rather than as a theorist. First, personnel: the party’s commanding heights were occupied, he said, by men born into the castes that had never suffered the disability, and their instincts travelled with them into the movement. Second, doctrine in application: caste was filed as an echo of property relations, to be cleared up after the main event, so the daily substance of untouchability, the well, the temple door, the dining line, went unaddressed. Third, practice: the unions took the workers as they stood, graded by birth inside the mill as outside it, and spent their solidarity on wages while sparing the hierarchy that divided the wage earners. His verdict followed: such a movement, victorious, would administer the old order in a new vocabulary.

Q: What is the argument in Annihilation of Caste?

The address written for a Lahore reform conference in 1936, and withheld from delivery when its hosts read it, maintains that the reformers had hold of the wrong end. Caste, it argues, is a system that ranks labourers themselves in a hereditary ladder, each rung taught reverence upward and contempt downward, and keeps the ladder standing by marrying strictly within each rung and by lending the whole arrangement the authority of sacred books. From those premises the conclusions fall in order: class unity among Indian workers is blocked at the source; political and economic reform alike will be ambushed by the doctrine that sanctifies the ladder; and nothing avails short of breaking the books’ authority, freeing marriage across the lines, and rebuilding religion around human equality.

Q: Why did Ambedkar prefer the Buddha to Marx?

On the evidence of the order the Buddha founded. Its members owned nothing privately, ate from a common round of alms, settled their affairs in assembly, and entered without anyone asking their birth, which demonstrated to his satisfaction that a community could practise common ownership on the strength of conviction alone. Generalising from the case, he held that transformations carried by persuasion lodge in the minds that must maintain them, while transformations carried by arms must be policed forever. The ends he largely granted to Marx; the ancient teacher seemed to him to have reached comparable ends by the one road that did not require a dictatorship to keep them reached, and the lecture awards the palm on exactly that ground.

Q: Was Ambedkar a socialist?

By the test of his own drafts, yes; by the test of party labels, of no existing school. The memorandum he laid before the constitution makers proposed a socialism of the commanding heights, land and great industry and insurance in public hands, fastened into fundamental law so that elections could administer but not repeal it. Around that economic core he insisted on wrappings the Marxist school discarded: a parliament answerable at fixed terms, courts that could void official acts, and a public morality, which he called religion, to police what law could not reach. Call the result democratic state socialism, with the accent distributed as he distributed it, on all three words at once.

Q: Do Marxists and Ambedkarites work together?

Often enough to disprove the legend of total separation, seldom enough to show why the legend persists. The nineteen thirties already contained a joint campaign against a Bombay labour bill that both movements judged a strikebreaker’s charter. The decisive attempts at fusion came later and from below: a Bombay organisation of young Dalit writers in the early nineteen seventies wrote class and caste into a single manifesto and a single street politics, a Maharashtra scholar rebuilt Marxist theory around Phule and Ambedkar as co founders, and an American born researcher spent a career writing the anti caste movement as a democratic revolution in Marxist terms. Against these stand the boundary’s keepers in both camps, for whom the other tradition remains an error to be corrected rather than a half to be completed.

Q: Where should you start with Ambedkar?

Start where he started in public, with the address the Lahore reformers would not hear. It is brief enough for a single sitting and contains, fully formed, the account of caste as a ranked division of the workers themselves, the attack on scriptural authority, and the standard of a society remade on equal terms. Follow it with the short lecture in which the founder of Buddhism and the author of Capital are weighed in one balance, reading the whole piece rather than the sentences each camp extracts. Then take the 1947 memorandum for the constructive scheme, land, industry, insurance, and the argument for entrenching them, and finish with the long scripture he composed for his converts, where the social teaching stands without the polemical frame.

Q: How does the partial acceptance grade his relation to Marxism?

It grades him element by element instead of awarding a single mark for the whole paper. Diagnosis of how exploitation works: accepted, and strengthened from Indian evidence. Objective of a society owning its productive wealth in common: accepted, and drafted. Account of reality as matter in dialectical motion: failed. Method of violent seizure followed by one class’s dictatorship: failed, the second failure entailed by doubts about the first. Estimate of religion as a dying illusion: failed, with a counter proposal attached. A grading of that shape explains two things a single mark cannot: why both camps can quote him honestly in their own support, and why neither can quote him completely without embarrassment.

Q: What should a student separate between accepting an end and rejecting the means proposed for it?

Separate the audit of destinations from the audit of vehicles, because they answer to different evidence. A destination is audited against the human goods it promises and the diagnosis that says those goods are blocked: here the questions are whether suffering really flows from property, and whether common ownership would really stop the flow. A vehicle is audited against its own physics: what it does to its passengers on the journey, what sort of driver it promotes, whether it can be parked when it arrives. Fluency in the first audit buys no competence in the second, and the habit of marking them on different lines is the single most transferable skill this debate teaches a student of political thought.

Q: Why did the debate turn on the means rather than the goal?

Agreement about the destination had removed the goal from contention. Both sides spoke of ending exploitation, breaking the power that ownership confers, and relieving the misery of the propertyless, in formulations close enough that each could borrow the other’s sentences. What remained was everything the agreement concealed: rival accounts of what moves history, of what the state is, of whether belief is a force or a shadow. Those accounts surface only when means are discussed, since a means is a philosophy with its sleeves rolled up. The debate therefore became a contest between a technology of consent, teaching, organisation, and entrenched law, and a technology of counter power, seizure, expropriation, and transitional rule, and that contest, not the shared destination, is where its permanent interest lies.

Q: How does the accepted and rejected ledger order his reading of Marx?

It converts a relationship usually described by temperature, warm or hostile, into a table that can be checked. Down one side run the components of the system; across run his verdict, the consideration that decided it, and the place where a sceptic can verify both. Reading by the ledger teaches a discipline that outlasts the topic: never ask whether a thinker was for or against a predecessor in the lump, since serious minds divide what they inherit, and the map of the division is the biography of their thought. In his case the map has a visible logic, welcome for whatever anatomised the present order, refusal for whatever proposed to govern men as material, and a student who can state that logic owns the debate rather than renting its slogans.

Q: What separates his criticism of the movement from his criticism of the doctrine?

The two criticisms live in different courts. Before the court of doctrine, the exhibits are books and the charges are errors: a metaphysics that cannot ground obligation, a transitional state that cannot be trusted to transition. The competent replies are better books, and the tradition produced them, in histories that gave caste material roots and in programmes that learned to name it. Before the court of the movement, the exhibits are rosters, manifestos, and the minutes of unions, and the charge is dereliction: that these men, in these years, let the hierarchy stand. The competent replies there are records, not arguments. Verdicts in one court bind nothing in the other, and most confusion about his position is a smuggling of evidence across the corridor between them.