Craft, Caste & Its Rejection

A world built on rank

To understand why the Sikh Gurus spoke so often about caste, you have to picture the society they lived in. In the Punjab of the fifteenth and sixteenth centuries, a person's place in life was largely decided at birth. Two overlapping systems were at work. One was varna ਵਰਨ, the old four-fold ranking of priests, warriors, traders, and labourers, with a group treated as so low it was placed outside the four. The other was jati ਜਾਤਿ, the much larger web of birth-groups tied to occupation, marriage, and region (Grewal 1998).

What caste decided

Caste was not only an idea. It shaped who you could marry, what work you could do, whose food you could eat, and whether you were even allowed to draw water from a shared well. The lowest groups, later called Dalit, were treated as ritually polluting. This is the ordinary, everyday cruelty the Gurus were addressing. It is important to be plain about it: caste was a system that told some human beings they were worth less than others from the moment they were born.

Why a scholar should be careful here

Historians warn against treating caste as one fixed, unchanging thing. It varied by place and shifted over time, and the British later hardened it through census categories (Singh and Fenech 2014). For this course, the key point is simpler: the Gurus inherited a society that ranked people by birth, and they said clearly that this ranking had no standing before the Divine. The rest of the course follows that claim from teaching, into institutions, and finally into the uncomfortable record of practice.

The central claim

The Sikh teaching on caste is not a footnote; it sits near the heart of the message. The Gurus taught that all people come from the same source and stand equal before the Creator. Birth into a high or low group tells you nothing about a person's true value. What matters is conduct and devotion, not the family one happened to be born into (McLeod 1997).

Caste as a spiritual error, not just a social problem

It would be too small to say the Gurus simply disliked caste as unfair. They went further and called the pride of caste a kind of self-delusion. To boast of one's birth-group is, in this teaching, to mistake an accident of birth for a real distinction, and to forget that everyone breathes the same breath. The Gurus pointed out that the same body, the same blood, and the same death come to all, so the labels of high and low are hollow.

Replacing the old measure

Crucially, the Gurus did not only criticise caste; they offered a different measure of a person. Worth is shown in honest living, remembrance of the Divine, and service to others. This reframing matters because a teaching that only says do not rank people leaves a vacuum. Sikh teaching fills it with a positive standard that anyone, of any background, can meet (Grewal 1998).

This is why the rejection of jati is not a side issue. It follows directly from the belief in one Creator present in every person. If that is true, no birth-group can be closer to or further from the Divine than another (McLeod 1976).

An idea made into a meal

Teachings can stay in the head. The Gurus built institutions so that equality would be acted out, not just believed. The clearest of these is langar ਲੰਗਰ, the free kitchen attached to the place of worship. Anyone, of any caste, religion, gender, or wealth, sits down and eats the same simple food, cooked and served by volunteers (Grewal 1998).

Why sitting matters

The detail that everyone sits together on the floor in a single line, the pangat ਪੰਗਤ, is not decoration. In a caste society, who you ate with, and at what level you sat, was a public statement of rank. Sharing food across caste lines was one of the strongest taboos. By making people of every background sit in the same row and eat from the same pot, langar broke that taboo on purpose, every single day. Tradition holds that even rulers who came to meet the Gurus were asked to sit in the pangat first.

The sociology of the shared kitchen

From a sociological angle, langar is a remarkable design. It does three things at once: it feeds the hungry, it trains the community in service, and it stages equality in a way the body remembers. A person who has eaten beside someone of another caste has crossed a line that mere words might never have moved. Scholars note that this institutional embedding is part of why the Sikh critique of caste was more durable than purely devotional movements that lacked such structures (Singh and Fenech 2014).

Langar is the answer to a fair objection: it is easy to say all are equal. The kitchen makes the community do it, in public, with their hands and their meals.

The gathered community

Alongside langar stands sangat ਸੰਗਤਿ, the congregation. In the sangat, people gather to sing and reflect together, and in principle no one's birth-group gives them authority over anyone else. The same scripture is open to all, the same floor seats everyone, and leadership in worship is not reserved for a priestly caste. This is a quiet but radical move: in the surrounding society, sacred knowledge was often guarded by one group (McLeod 1976).

The dignity of craft

This course is titled craft and caste for a reason. In the caste system, occupation and birth-group were welded together; your craft was your rank, and a lowly craft made you a lowly person. Sikh teaching cut that link. Honest work, kirat ਕਿਰਤਿ, is dignified in itself. The Gurus praised earning an honest living and sharing it, and they did not treat manual or artisan labour as shameful. The point was not that craft does not matter, but that no craft makes one human worth more or less than another.

Institutions that carry values

Put langar, sangat, and the honouring of honest work together and you can see a deliberate strategy. The Gurus did not rely on people simply agreeing with a teaching. They embedded the teaching in repeated practices: where you sit, who you eat with, who may lead worship, and how work is spoken of. Sociologists call this the institutionalisation of values, and it is a large part of why the Sikh rejection of caste became a recognisable feature of the community rather than a passing sermon (Grewal 1998).

So far the course has shown the ideal and its institutions at their best. The next lesson turns to the harder question every honest student must ask: did it work?

Ideal and reality are not the same thing

A serious course cannot stop at the ideal. The sociological record is clear that caste did not disappear among Sikhs, even where the teaching plainly condemns it. To pretend otherwise would be dishonest and would insult the people who have suffered from caste prejudice within the community. So this lesson holds the ideal up against lived practice, as historians and sociologists describe it (Jodhka 2014).

Where caste survived

The most stubborn site of caste is marriage. Many Sikh families still arrange marriages within their biradari ਬਰਾਦਰੀ or caste-group, so that named identities such as Jat, Khatri, Arora, Ramgarhia, and others have stayed socially real across generations. Caste also showed up in the control of gurdwaras, in landholding and rural power, and at its most painful, in the lower social standing given to Sikhs of Dalit background, such as those called Mazhabi ਮਜ੍ਹਬੀ or those associated with Ravidasi traditions. In a number of villages, separate gurdwaras or separate cremation grounds emerged along caste lines (Jodhka 2014).

Why did it persist?

Several forces worked against the ideal. Sikh society remained embedded in the wider Punjabi and South Asian social order, where caste organised land, kinship, and politics. Conversion to a teaching does not instantly erase centuries of inherited habit, especially around marriage and inheritance. Scholars also point to the way British colonial administration recorded and rewarded caste categories, and to the simple fact that dominant land-owning groups had every worldly reason to keep their advantages (Singh and Fenech 2014). None of this excuses the gap; it explains how a community can sincerely hold a teaching and still fall short of it.

The honest conclusion is uncomfortable but clear: the Sikh tradition rejected caste in principle more decisively than most of its neighbours, yet caste survived in Sikh practice, and in some forms it survives still.

Two truths at once

This course has asked you to hold two things together. First, that the Sikh Gurus rejected caste clearly and built institutions, langar and sangat, that act out the equal worth of all. Second, that caste nonetheless persisted among Sikhs in marriage, gurdwara life, and the treatment of Dalit Sikhs. A mature reader does not collapse one truth into the other. The teaching is not cancelled by the failure to live it, and the failure is not erased by the beauty of the teaching (Grewal 1998).

How scholars frame the gap

Historians of religion are used to this pattern. Almost every tradition holds ideals its followers do not fully reach. The interesting question is what a tradition does with its own standard. In the Sikh case, the standard itself, equality before the Divine, gives reformers a tool: caste discrimination can be condemned from inside the tradition, on the tradition's own terms, rather than only from outside (McLeod 1997). Movements and individuals within the Panth have repeatedly used the teaching to challenge caste practice.

Why honesty serves the ideal

It might seem that admitting the persistence of caste weakens the Sikh claim. The opposite is true. A tradition that can name its own gap between teaching and practice is more credible, not less, and better placed to close that gap. Pretending caste vanished would leave its victims unseen and its critics unanswered. Naming it honestly keeps the original teaching alive as a demand, not just a slogan (Jodhka 2014).

The takeaway

The Sikh rejection of caste is real, principled, and embedded in daily practice through langar and sangat. The persistence of caste among Sikhs is also real. Studying both, without flinching from either, is what it means to take the tradition seriously as both a teaching and a living community (Singh and Fenech 2014).