There is a dish in Puducherry that nobody outside the town makes: a fish curry in coconut milk, flavoured with curry leaf and aniseed, finished with a squeeze of lime. It is served with rice, and in the houses where it is still cooked it may also be served with a piece of baguette. The curry is not French. The bread is not Tamil. Both are from the same kitchen, and that kitchen has been running for about two hundred and eighty years.
What makes this worth a chapter is not that a colonial cuisine exists — they exist everywhere. It is that the crossover here happened at a level most fusion food never reaches. Almost nothing in this cooking is a hybrid ingredient. There is no curry pizza, no tandoori quiche. What was mixed was the order of operations, and that is why you can eat the result without noticing that anything was mixed at all.
Two towns in one
The French arrived in 1674 and stayed until 1954, when the town voted to join India; the treaty took another eight years to finish. For most of that time there were two administrations, two legal systems, two school systems and, in the same streets, two kitchens. A French colonial survey published in 1889 is revealing about how separate they were: it gave its pages to what it called the European diet, and disposed of the rest in a single clause — "leaving aside the food of the Indian class". The two cuisines were not being blended by anybody who wrote things down. They were being recorded as two facts about one town.
So the mixing happened where records are not kept: in households. Specifically in households that were French and Tamil at once — the families of French men who settled, married or otherwise formed households locally, and whose daughters and cooks ran a kitchen that had to feed people who expected bread and people who expected rice. That is the usual engine of creole cooking everywhere: not a chef, a household, and a mother who is not interested in theory.
Why these two cuisines could merge
Here is the structural reason, and it is the whole argument of this piece. Both the French kitchen and the Tamil kitchen begin the same way: fat in a pan, aromatics into the fat, then liquid, then a long simmer. The French version uses butter, onion, carrot and celery; the Tamil version uses oil, mustard seed, curry leaf and dried chilli. They are not the same ingredients, but they are the same move — and because the grammar matches, every element after it can be substituted one for one.
Every substitution in the middle of that diagram is the same trade. Wine becomes tamarind water, because both are the acid that cuts fat and both are what the place has. Stock becomes coconut milk, because both carry body and both are cheap locally. Cream becomes the first press of the coconut, because both are fat and both thicken. Aniseed comes the other way, from the European side, and it is the tell: a Tamil fish curry does not normally contain aniseed. When you taste it, you are tasting the seam.
The artefact that proves it
If the argument is that the crossover was technical, there should be an object that shows it. There is: vadouvan. Take onion, garlic, curry leaf and mustard seed, grind them together, roll the paste into balls and leave them in the sun for weeks. What you get is the first two rows of that diagram in a form that will sit in a jar for a year — a mirepoix that keeps, and that gets stronger rather than weaker as it ages.
Vadouvan is not a colonial invention; it belongs to Tamil cooking well beyond this town, and every household that uses it has its own proportion. What is specific to Puducherry is what it was used to start: a French braise. A cook who wants to begin a dish the French way but season it the Tamil way has exactly one problem — the aromatics are fresh and perishable, and the French method assumes you have them on the day. The ball solves it. That is what a real fusion artefact looks like: not a dish that tastes of two places, but a technique from one place rebuilt to fit the constraints of the other.
Three layers, not two
The dishes carry a third layer that is easy to miss. Fish assad takes its name from the Portuguese assado, not from anything French. Vindail — a sour, garlicky meat curry — is the town's cousin of Goa's vindalho, but it is soured with tamarind rather than vinegar and it has no potato, so it is a different solution to the same Portuguese problem. The Portuguese were in the Indian Ocean a century before the French arrived, and their words stuck to the techniques while their administration left. A town can have three kitchens in its vocabulary and only remember two.
And then there is the bread. In older households the everyday loaf was called roll roti, and was dipped in sweet tea or in coconut milk; the croissant was called erral roti — prawn bread — because of its curve. That is the most honest evidence in the whole story. Nobody invented a word for a croissant. The town named it after something it already had, which is what people do with a new object they intend to keep: they fit it into the language they already speak. The dish changed less than the vocabulary did.
Why it is disappearing from menus
This cooking is now hard to find in restaurants and survives mostly in homes, taught by older women in classes that are partly a business and partly a rescue. The reason is economic and slightly sad. A creole kitchen is, by definition, double: two pantry systems, two sets of skills, two shopping lists. A restaurant optimises for one, because the second system is dead stock. A home does not optimise; a home just cooks what the family eats, and if the family eats both, both stay alive.
There is a broader point here about fusion as a genre. The fusions that survive are the ones where the grammars matched — where the first move was the same, so the substitutions could be made without breaking the dish. The fusions that die on the plate are the ones that only swap ingredients: a curry on a pizza, a tandoori filling in a croissant. Those are juxtapositions, and juxtaposition is a menu idea, not a cuisine. What happened in Puducherry took two hundred and eighty years, no chef, and nobody calling it fusion. It is the only kind that lasts.
Sources and checks
- Henrique, Louis, ed. Les colonies françaises: notices illustrées, vol. 2, L'Inde française. Paris: Maison Quantin, 1889
- Miles, William F. S. Imperial Burdens: Countercolonialism in Former French India. Lynne Rienner, 1995
- Achaya, K. T. A Historical Dictionary of Indian Food. Oxford University Press, 1998
- Sen, Colleen Taylor Feasts and Fasts: A History of Food in India. Reaktion Books, 2015
- Gracias, Fatima da Silva Cozinha de Goa: History and Tradition of Goan Food. 2nd ed., 2019 — for the vindalho family
- Banerjee, Poulomi Crepes to curries: the changing tastes of Pondicherry. Hindustan Times, 28 April 2019
Last checked 2 November 2026 Yellowsea Huang Corrections