More than eight thousand years ago, a remarkable thing happened that changed the course of a fruit and mankind, at least in a certain latitude to start with. A few wild vines bearing small berries, through a random series of reshufflings in their chromosomes, became hermaphrodites [1, 2, 3]. Hermaphrodite is a living being which loses its sex denomination by fusing both sexes within one. Hence it can pollinate itself and bear fruit, and seed, without needing a partner.

The human population were possibly tending to the wild vines before this switch happened [1]. But after this switch, it was a paradigm shift as the wasted space for the male vine was now replaced by a fruit bearer. This likely propelled the large scale adoption of this vine and eventually wine production.

But we need to be aware that from early on, these vines were most likely propagated from cuttings. That much our ancestors had figured out. Seed never gave back the vine you loved, so people planted pieces of the vine itself [5].

But I have always had these questions that intrigued me: why grape and why wine? Let us first sit with the question of the fruit. People would have tried every fruit that grew in that belt. The question was not so much what people liked as what survived the jar. The filter was microbes first, taste second.

There were also certain intrinsic qualities that made grape a winner. The first of those, the one everything else depended on, was to have enough sugar to make alcohol strong enough to keep, about 13 to 14 percent unaided, where apple juice stops at cider strength [6, 7, 8].

The second and most consequential of all properties is that grape has tartaric acid, which is an acid that fermentation cannot eat. Apple, peach, cherry, pear and plum all run on malic acid and berries like blueberry, raspberry, blackberry, cranberry, currants run on citric acid [6]. Tartaric acid is uncommon in fruit, and among the fruits that grew in that belt, only the grape stores it in large amounts. A ripe grape carries 4.5 to 10 grams of it per liter, against 2 to 6.5 grams of malic acid [9]. Malic acid is used up for energy during the maturation of the fruit while it builds up sugar and in the process gets diluted. Tartaric acid, though, stays largely untouched; the vine makes it early and keeps it through ripening [10, 11]. The bacteria that work in wine can all eat malic acid, but neither they nor the fermenting yeast can use tartaric acid [9, 11]. Once the grape is crushed, the acid holds back the microbes that spoil wine and lets the yeast through to make it [12]. So, when apples ferment, the acid itself becomes food. But when a grape ferments, the acid becomes a wall. Together with the alcohol, this lets wine last far longer in the jar.

So, when apples ferment, the acid itself becomes food. But when a grape ferments, the acid becomes a wall.

The third test that grape passes is that it comes with the sugar that is most ready to ferment. The leaves make sucrose from the sun and pass it onto the berries. An enzyme in the fruit splits it into nearly equal parts of glucose and fructose [13]. For example, ripe Pinot noir flesh is almost all simple sugar, with 84 milligrams of glucose and 81 milligrams of fructose per gram [13]. This matters because the fruit hands the yeast food it can eat as it is. When grain is used to make alcohol, it needs malting or mold to break starch to sugar before yeast can ferment [14]. For other fruits, we have to add sugar to reach a strength that keeps [15].

The fourth advantage is that grapes have thin skins, grow in tight bunches and are mostly water. So if you pile grapes in a container, the bottom layer would burst under the weight of the top without a need for a press [16]. So wine, unlike any other fruit, does not need to be made, only noticed.

So wine, unlike any other fruit, does not need to be made, only noticed.

Next we have the reason for the grip wine leaves in the mouth, the tannins. They come from the grape’s skins and seeds. These tannins react with oxygen, taking on damage the wine would otherwise take [17]. This is one of the reasons why wine lasts for years, not months, especially the red ones as they are fermented with the skins [18].

Tannin is a chain of ring-shaped molecules, each studded with hydroxyl (OH) groups. The building blocks of this substance are found in tea and cocoa [19]. Skin tannins come in longer chains than seed tannins, and red wine has more of both because it ferments with them [17, 18]. The hydroxyl groups help the tannin latch onto proteins in the saliva, which we feel as grip in the mouth, just like when the rubber meets the road [20]. These chemicals have long been used to do the same to the proteins in animal skin, turning hide into leather. That is where tannin gets its name [21]. These hydroxyl groups also love oxygen, reacting with it readily and so protecting the wine.

The last advantage the grape carries is what gives wine its character. These are smells locked deep inside the grape, unlocked by fermentation and time to be finally sensed by us [22, 23]. Every wine may seem to run the same script, fermentation, but the cast of molecules each grape brings decides how the film plays.

Now, we turn to the question we wanted to answer earlier. Why wine? The fruit itself answers that question fairly well. But there is more to it than that.

The vine itself has plenty of varieties, and every vine of a variety is a cutting from one original plant, a copy that can be passed on reliably for centuries. The story of varieties is fascinating as these vines have no sex denomination and have the ability to pollinate themselves. The varieties still emerge in two ways. Just because they pollinate themselves does not mean that the pollen may not drift into a field with a different vine altogether. This is likely what happened with Cabernet sauvignon, when a chance cross-pollination between Cabernet franc and Sauvignon blanc took place in Bordeaux, probably in the 1600s, and every Cabernet since is a copy of the vine grown from that seed [24]. The other way you get a variety is called a bud sport, which scientists call a somatic mutation. In this case a single shoot on an existing vine changes on its own and cuttings from that shoot carry the change forward for good. No seed, no new parents, the same plant changed in one way on that one branch. Pinot gris and Pinot blanc arose like this from Pinot noir [25]. This is what makes the grape and its wine very interesting indeed.

Now that we know how each variety came to be, here comes the cast that plays fermentation’s script, each variety’s own molecules, such as Syrah’s pepper and Sauvignon blanc’s grapefruit. Syrah’s pepper is rotundone, the same molecule found in black peppercorns [26]. Sauvignon blanc’s grapefruit is a thiol that has no smell at all in the grape until the yeast sets it free [27]. The actors in this cast are ones we already know from real life, in other plants and in the things we use.

The rest of the cast is just as familiar. Muscat and Riesling carry linalool and geraniol, the floral molecules we know from lavender and roses, and much of their perfume sits locked to sugars in the grape until fermentation sets it loose [28, 29, 30]. Gewürztraminer brings rose oxide, the molecule behind its smell of lychee [31]. Cabernet sauvignon carries a methoxypyrazine, the green smell of a bell pepper, which fades when sun reaches the grapes early in the season [32, 33]. And one actor arrives late. In an aging Riesling, pigments from the grape slowly turn into TDN in the bottle, a whiff of kerosene that grows stronger with time [34].

Then there are some grapes that bring a small cast, a faint one at that, and those lean heavily on fermentation and the winemaker [28]. These are only a few of the actors. Every grape variety and the molecules that make it what it is are being gathered in the Terroir and the Fruit in the Glass database.

We now move from the vine and the fruit to the leaves, which feed them both.

These leaves are the solar panels above the vine. Light from our sun, eight minutes away, lands inside tiny green compartments in the leaf cells called chloroplasts, where thousands of chlorophyll molecules, yes the ones we heard about in class, sit on thin membranes in light-catching teams [37, 38]. The one that catches the light passes its energy along like a baton in a relay until it reaches the team’s center. There it kicks an electron loose. The chlorophyll, now short one electron, is so hungry for it that it rips electrons out of water, with the help of a tiny cluster of manganese and calcium [39, 40]. The water falls apart. The freed electrons spend their energy pumping hydrogen ions to one side of the membrane, where they pile up like water behind a dam. Their only way out is a spinning turbine, and every turn makes ATP, the energy coin of all known life [41]. A second burst of light loads the electrons and hydrogen onto a molecular truck, and with the ATP they turn carbon dioxide into sugar [38]. The hydrogen in the grape’s sugar was pulled out of water by starlight. The carbon came from the air, and the leaf breathes the water’s oxygen out [42].

The hydrogen in the grape’s sugar was pulled out of water by starlight.

The leaves use their pores as the molecular exchange doorway letting carbon dioxide in and water out [43]. The leaf’s tank of sunlight is full at about a third of full sun, and more light than that adds nothing [44, 45]. In theory at most 4.6 percent of the light that hits the leaves can be converted to sugar which is a low energy conversion rate in itself but in a real vineyard, it is even lower [46]. Essentially a dozen leaves feed a grape bunch [47]. The heat from the sun, though, slows the sugar-making above 35°C (95°F), one more reason the vine does best within the temperature window we met in Part 1 [48]. A lot of the leaves are hidden from the sun and do almost no sugar-making [45, 49].

Traveling beneath, we reach the roots the vine stands on, where the water the leaf split comes up from the soil.

Vine roots can reach past 6 meters, deeper than many plants, yet most of them sit in the top meter of soil [50]. The leaves are the ones pulling most of the water as one continuous chain as the water leaves them, not the roots [51]. The one time the roots push is early spring, before any leaves open, when a freshly pruned vine drips sap from the cut [52]. The starch in the root feeds the first spring leaves acting as a pantry for the vine [53, 54]. Most vines today stand on the roots of a different plant [55]. Phylloxera, an insect from North America, killed most of the European vineyards in the late 1800s. The fix was grafting European vines onto American roots, which had grown up alongside the insect and could live with it [56]. The irony is that Carménère, a grape that grafted poorly and all but vanished from Bordeaux, survived on its own roots in Chile, which the insect never reached [57]. Those Chilean vines are a living link to the Bordeaux of the 1800s. Yet that wine cannot carry Bordeaux’s name or its prestige, because the name belongs to the place, not to the plant.

The name belongs to the place, not to the plant.

The roots do not work alone. Fungi wrap around them and thread into the soil far past where the root can reach, bringing back water and nutrients like phosphorus and helping fend off root disease, and the vine pays them in sugar [58]. And for those who taste the rocks in their wine, the vine does not carry them up. The minerals that reach the grape are dissolved nutrients in tiny amounts, with no taste of their own [59].

People have been deciding which plants belong where for a long time. On 31 July 1395, Philip the Bold, Duke of Burgundy, ordered every Gamay vine in his lands cut down within a month and torn out by the next Easter. His edict called Gamay “a very bad and very disloyal plant” and its wine “very harmful to humans,” and it set fines, with a quarter paid to informers. The same decree scolded growers for leaving good vineyard land in ruin, for feeding Gamay with manure and for adding warm water to its wine. Pinot noir kept Burgundy’s hillsides. Gamay already grew in Beaujolais, just to the south, where it thrives today. A human decided which plant could grow in which place, more than six hundred years ago, by law [60].

The same thing happened again in the twentieth century, in Georgia. The country counts around 525 grape varieties of its own [61, 62]. Under Soviet rule, wine became a matter of volume, and high-yielding Rkatsiteli and Saperavi came to cover most of its vineyards [62]. Only a handful of varieties made it into wine for sale [63]. The rest survived because families were allowed small private plots, and there they kept their grandparents’ vines alive [63]. This time a state decided which plants would grow, and backyards saved the others.

And so we end with the acid the fermentation cannot eat. Eight thousand years on, it is still there. In clay jars dug up at two Neolithic villages in Georgia, Gadachrili Gora and Shulaveris Gora, chemists found tartaric acid soaked into the pottery, from about 6000 to 5800 BC [64]. At Hajji Firuz in the mountains of Iran, they found it in jars from about 5400 to 5000 BC [65]. They look for this one molecule because, of all the fruits people traditionally turned into fermented drinks, only the grape carries it in large amounts [16]. The wall that kept the wine from spoiling is also the fingerprint that tells us it was there. This fruit and its wine have earned every claim to being special, and none of it is imagined.

Next in the series, we follow the wine out of the vineyard and into human hands.

Sources

1. Massonnet, M. et al. (2020). The genetic basis of sex determination in grapes. Nature Communications 11, 2902.

2. Dong, Y. et al. (2023). Dual domestications and origin of traits in grapevine evolution. Science 379, 892 to 901.

3. Zou, C., Massonnet, M., Minio, A. et al. (2021). Multiple independent recombinations led to hermaphroditism in grapevine. PNAS 118(15), e2023548118. doi:10.1073/pnas.2023548118

4. Coito, J. L. et al. (2019). Vitis flower types: from the wild to crop plants. PeerJ 7, e7879.

5. Myles, S., Boyko, A. R., Owens, C. L. et al. (2011). Genetic structure and domestication history of the grape. PNAS 108(9), 3530 to 3535. doi:10.1073/pnas.1009363108

6. Fiola, J. Fruit Wine Chemistry and Fermentation. University of Maryland Extension.

7. Australian Wine Research Institute (2016). Predicting alcohol levels. Grapegrower and Winemaker 626.

8. Peck, G., Miles, C., King, J., Bradshaw, T., Rothwell, N. and Merwin, I. (2019). An Introduction to Hard Cider in the U.S. eXtension, Apples community.

9. Mendes Ferreira, A. and Mendes-Faia, A. (2020). The role of yeasts and lactic acid bacteria on the metabolism of organic acids during winemaking. Foods 9(9), 1231. doi:10.3390/foods9091231

10. DeBolt, S., Cook, D. R. and Ford, C. M. (2006). L-tartaric acid synthesis from vitamin C in higher plants. PNAS 103, 5608 to 5613.

11. Ribéreau-Gayon, P. et al. (2006). Handbook of Enology, Volume 2, 2nd ed., chapter 1, Organic acids in wine. Wiley.

12. Zoecklein, B. Controlling microbial growth in wine. Virginia Tech.

13. Walker, R. P., Bonghi, C., Varotto, S. et al. (2021). Sucrose metabolism and transport in grapevines, with emphasis on berries and leaves, and insights gained from a cross-species comparison. Int. J. Mol. Sci. 22(15), 7794. doi:10.3390/ijms22157794

14. Parés Viader, R. et al. (2021). Optimization of beer brewing by monitoring α-amylase and β-amylase activities during mashing. Beverages 7(1), 13; and Takeshita, N. (2020). Fungal research in Japan: tradition and future. Fungal Biology and Biotechnology 7, 14.

15. North Dakota State University Extension (2021). FN1638, making fruit wines.

16. McGovern, P. E., Hartung, U., Badler, V. R., Glusker, D. L. and Exner, L. J. (1997). The beginnings of winemaking and viniculture in the ancient Near East and Egypt. Expedition 39(1). Penn Museum.

17. Jeremic, J., Vongluanngam, I., Ricci, A., Parpinello, G. P. and Versari, A. (2020). The oxygen consumption kinetics of commercial oenological tannins in model wine solution and Chianti red wine. Molecules 25(5), 1215. doi:10.3390/molecules25051215

18. Harbertson, J. Grape and Wine Phenolics: A Primer. Washington State University Viticulture and Enology.

19. Haytowitz, D., Wu, X. and Bhagwat, S. (2018). USDA Database for the Proanthocyanidin Content of Selected Foods, Release 2.1. USDA Agricultural Research Service; Gu, L., Kelm, M. A., Hammerstone, J. F. et al. (2004). Concentrations of proanthocyanidins in common foods and estimations of normal consumption. J. Nutr. 134(3), 613 to 617; Shi, Q.-Y. and Schlegel, V. (2012). Green tea as an agricultural based health promoting food, the past five to ten years. Agriculture 2(4), 393 to 413. doi:10.3390/agriculture2040393.

20. Kuhlman, B. A., Aleixandre-Tudo, J. L., Moore, J. P. and du Toit, W. (2024). Astringency perception in a red wine context, a review. OENO One 58(1). doi:10.20870/oeno-one.2024.58.1.7774

21. Online Etymology Dictionary, “tannin.”

22. Doneva-Šapeska, D. et al. (2006). Bulletin of the Chemists and Technologists of Macedonia 25, 51 to 56.

23. Marais, J. (1987). Terpene concentrations and wine quality of Vitis vinifera L. cv. Gewürztraminer as affected by grape maturity and cellar practices. Vitis 26, 231 to 245.

24. Bowers, J. E. and Meredith, C. P. (1997). The parentage of a classic wine grape, Cabernet Sauvignon. Nature Genetics 16, 84 to 87.

25. Vezzulli, S., Migliaro, D. et al. (2012). Pinot blanc and Pinot gris arose as independent somatic mutations of Pinot noir. Journal of Experimental Botany 63, 6359.

26. Wood, C., Siebert, T. E., Parker, M. et al. (2008). From wine to pepper, rotundone, an obscure sesquiterpene, is a potent spicy aroma compound. J. Agric. Food Chem. 56(10), 3738 to 3744. doi:10.1021/jf800183k

27. Roland, A., Schneider, R., Razungles, A. and Cavelier, F. (2011). Varietal thiols in wine. Chemical Reviews 111, 7355 to 7376.

28. Mateo, J. J. and Jiménez, M. (2000). Monoterpenes in grape juice and wines. J. Chromatogr. A 881, 557 to 567.

29. Pokajewicz, K. et al. (2021). Molecules 26, 5681 (lavender oil composition).

30. Dobreva, A., Nedeva, D. and Mileva, M. (2023). Resources 12, 83 (rose oil composition).

31. Chigo-Hernandez, M. M., DuBois, A. and Tomasino, E. (2022). Aroma perception of rose oxide, linalool and α-terpineol combinations in Gewürztraminer wine. Fermentation 8(1), 30.

32. Harris, R. L. N. et al. (1987). Vitis 26, 201 to 207 (2-isobutyl-3-methoxypyrazine, the bell pepper aroma).

33. Ryona, I., Pan, B. S., Intrigliolo, D. S., Lakso, A. N. and Sacks, G. L. (2008). Effects of cluster light exposure on 3-isobutyl-2-methoxypyrazine accumulation and degradation patterns in red wine grapes (Vitis vinifera L. cv. Cabernet Franc). J. Agric. Food Chem. 56, 10838 to 10846. doi:10.1021/jf801877y

34. Tarasov, A., Giuliani, N., Dobrydnev, A., Schuessler, C., Volovenko, Y., Rauhut, D. and Jung, R. (2020). 1,1,6-Trimethyl-1,2-dihydronaphthalene (TDN) sensory thresholds in Riesling wine. Foods 9(5), 606.

35. International Organisation of Vine and Wine (2017). Distribution of the World’s Grapevine Varieties.

36. Romero, R., Chacón, J. L., García, E. and Martínez, J. Pyrazine contents in four red grape varieties cultivated in warm climate. OENO One (formerly J. Int. Sci. Vigne Vin), article 861.

37. Smithsonian National Air and Space Museum. How long does it take light from the Sun to get to Earth?

38. Hardy, College Biology I, section 24.3, adapted from OpenStax Biology 2e, CC BY 4.0.

39. Merchant, S. and Sawaya, M. R. (2005). The light reactions. Plant Cell 17, 648 to 663.

40. Umena, Y., Kawakami, K., Shen, J.-R. and Kamiya, N. (2011). Crystal structure of oxygen-evolving photosystem II at a resolution of 1.9 Å. Nature 473(7345), 55 to 60. doi:10.1038/nature09913

41. Royal Swedish Academy of Sciences (1997). Press release, Nobel Prize in Chemistry (ATP synthase).

42. Dorough, G. D. and Calvin, M. (1951). The path of oxygen in photosynthesis. J. Am. Chem. Soc. 73, 2362.

43. Morris, J. et al. Biology: How Life Works, chapter 29.5.

44. Robinson, J. (ed.). The Oxford Companion to Wine, entry “photosynthesis.”

45. Smart, R. (2019). How much shade is too much? IWC Insight.

46. Zhu, X.-G., Long, S. P. and Ort, D. R. (2008). What is the maximum efficiency with which photosynthesis can convert solar energy into biomass? Current Opinion in Biotechnology 19(2), 153 to 159. doi:10.1016/j.copbio.2008.02.004

47. Kliewer, W. M. and Dokoozlian, N. K. (2005). Leaf area/crop weight ratios of grapevines. Am. J. Enol. Vitic. 56, 170 to 181.

48. Luo, H.-B., Ma, L., Xi, H.-F. et al. (2011). Photosynthetic responses to heat treatments at different temperatures and following recovery in grapevine (Vitis amurensis L.) leaves. PLoS ONE 6(8), e23033. doi:10.1371/journal.pone.0023033

49. Wolf, T. K. Canopy management, chapter 7, North Carolina Winegrape Grower’s Guide. NC State Extension, 2007, revised 2025.

50. Smart, D. R., Schwass, E., Lakso, A. and Morano, L. (2006). Grapevine rooting patterns, a comprehensive analysis and a review. Am. J. Enol. Vitic. 57(1), 89 to 104, as summarized in the UC Davis Viticulture and Enology research summary.

51. Venturas, M. D., Sperry, J. S. and Hacke, U. G. (2017). Plant xylem hydraulics. J. Integr. Plant Biol. 59, 356 to 389.

52. Stafne, E. Grapevine bleeding. eXtension.

53. Pellegrino, A. et al. (2014). Frontiers in Plant Science 5, 283 (root starch).

54. Zufferey, V., Murisier, F., Vivin, P. et al. (2012). Carbohydrate reserves in grapevine (Vitis vinifera L. Chasselas), the influence of the leaf to fruit ratio. Vitis 51(3), 103 to 110.

55. Ollat, N., Bordenave, L., Tandonnet, J. P., Boursiquot, J. M. and Marguerit, E. (2016). Grapevine rootstocks, origins and perspectives. Acta Horticulturae 1136, 11 to 22. doi:10.17660/ActaHortic.2016.1136.2

56. Skinkis, P., Walton, V. and Kaiser, C. (2009). Grape Phylloxera: Biology and Management in the Pacific Northwest. Oregon State University Extension EC 1463-E.

57. Pszczólkowski, Ph. (2013). Carmenère, mayoría de edad, 1994 to 2012. Estudios Avanzados 20, 137 to 153; Botton, M., Ringenberg, R. and Zanardi Zanuzo, O. (2004). Ciência Rural 34(5), 1327 to 1331.

58. Trouvelot, S. et al. (2015). Arbuscular mycorrhiza symbiosis in viticulture: a review. Agronomy for Sustainable Development 35, 1449 to 1467.

59. Maltman, A. (2013). Minerality in wine: a geological perspective. Journal of Wine Research.

60. Phillips, R. (2023, September 21). By decree, the Duke of Burgundy and a 14th-century AOC. World of Fine Wine.

61. Maghradze, D., Failla, O., Bacilieri, R. et al. (2019). Georgian Vitis germplasm, conservation, research and usage. 21st GiESCO International Meeting, Thessaloniki, 130 to 131.

62. Mazzeo, J. (2022, October 7). Georgia’s indigenous grapes, reviving hidden treasures. Decanter.

63. Alfonsi, S., Chasan, A., Velie, A. and Dozor, J. (2023, November 26). In Georgia, wine’s surprising birthplace, winemakers revive ancient grapes. CBS News, 60 Minutes.

64. McGovern, P., Jalabadze, M., Batiuk, S. et al. (2017). Early Neolithic wine of Georgia in the South Caucasus. PNAS 114(48), E10309 to E10318. doi:10.1073/pnas.1714728114

65. McGovern, P. E., Glusker, D. L., Exner, L. J. and Voigt, M. M. (1996). Neolithic resinated wine. Nature 381, 480 to 481. doi:10.1038/381480a0

    Subscribe now

Below this line is the first edition of the Terroir and the Fruit in the Glass database. It covers 13 grapes, from Cabernet Sauvignon to Trebbiano, with the molecule behind each smell, each one drawn as its chemical structure and traced to the research that found it. Paid subscribers can read it here and download it as a PDF. More varieties are being added.

Read more

This is Part 2 of The Physics of Travel. Part 1 covered sunlight, dew point, sweat and the body.

Besides all the factors we discussed in Part 1, something systemic has been at play, one that has shaped many people’s travel and certainly mine. I have traveled to Chennai, India, all my life, mostly in late July and August. Lately I have been chasing better conditions by going later and later, and now even October is hot. The same has happened with Boston and New York, where we go to escape the Phoenix summer; those trips, too, have slid later. And our longer vacations, to Barcelona, Rome, and now Marrakech, have backfired.

The records show where that feeling is right, and where it is not. I compared the same two stretches of fifteen years for every city, 1991 to 2005 and 2011 to 2025, using each city’s official airport record of daily highs (NOAA GHCN-Daily [1], which has gaps in some recent years; for Marrakech, the airport’s own reports via the Iowa Environmental Mesonet [2]). For each place, I called a day hot if it beat the hottest tenth of days in the earlier years, so every city is judged against its own past.

Chennai changed in two ways. At Meenambakkam airport, July and August barely warmed by the thermometer, about a degree Fahrenheit on the average high. But the air grew stickier. The dew point rose about two degrees, to 75, and the heat index, the temperature the humid air feels like, rose about four degrees, to an average daily high of 107 [2]. On skin at 93 degrees, that cut the pull that evaporates sweat by about 7 percent. So the months I once found bearable became harder to bear, not because they got much hotter but because sweat stopped working as well. October changed by the thermometer itself. Its average high rose almost three degrees, from 90 to 93, and its hot days, those above 94 degrees, went from about one in eleven to one in three. October, the month I moved my visits to, is the month that has gone.

Our summer escapes tell a different story. At LaGuardia, New York’s August has barely moved: its average high rose half a degree, to about 85, with no more hot days than before. Boston’s August warmed a little more, about a degree. What has always been true is the gap. Late September in New York runs about nine degrees cooler than its August, and in Boston about ten. That cool window is why our trips slid there, and it still holds, though it too has warmed by one to two degrees.

Even London, a city I never thought of as warm, has moved. In the Met Office record at Heathrow, the average September high rose from about 68 to almost 70 degrees between the same two periods, about as much as its summer warmed [3].

We tried Europe too, on the advice that Barcelona and Rome were not too hot in high summer. That advice described a climate that has since shifted. At Barcelona’s El Prat airport, the average July high rose two and a half degrees, to about 84, and its hot days, above 87, rose by about three-quarters. At Rome’s Ciampino airport, the July average high rose three degrees, to about 90, and its hot days, above 92.5, more than tripled, from one in ten to almost one in three. Hot trips like ours have become far more likely than they were.

Marrakech this September fits the same pattern. At Menara airport, the September average high rose almost three degrees between the two periods, and its hot days, above 100, rose by about two-fifths. Then came the September we walked into. Its average high was 98 degrees, more than two degrees above any other September in the airport’s reports since 1990, with 24 days of 95 or more and 16 of 100 or more [2].

The average creeps up slowly, but the hot days multiply fast. In Chennai’s October, a rise of less than three degrees in the average high came with almost four times as many hot days. That is the arithmetic of a warming climate. When the whole range of temperatures shifts even slightly upward, the days at the hot end, the ones that ruin a walk or a trip, increase far faster than the average.

Why is this happening? The tilt is not changing on any timescale that matters to us. Over the last million years, it has varied between 22.1 and 24.5 degrees, in a cycle of about 41,000 years [4]. What climate change alters is the Earth’s energy budget. Carbon dioxide absorbs the infrared heat the Earth radiates upward, in a few bands of the infrared, and sends part of it back down, a cousin of the physics of a cloudy night, but everywhere and around the clock. Its concentration has risen from about 280 parts per million before industrialization to about 428 today, higher than at any time in at least two million years [5, 6]. That adds roughly 2.3 extra watts to every square meter of the planet from carbon dioxide alone, averaged over the globe and the year [7]. The Earth’s surface over the last decade has been about 1.3 degrees Celsius, or 2.3 degrees Fahrenheit, warmer than in the late 1800s [8], and hot extremes have become more frequent and more intense across most of the world’s land since the 1950s [6].

Warming also raises the saturation vapor pressure, about 7 percent for every degree Celsius [9], the curve in Figure 3, in Part 1, so where moisture is available a warming world trends toward higher dew points and heavier downpours. And the seasons themselves have stretched. In the Northern Hemisphere’s middle latitudes, summer, counted as the stretch of the year’s hottest temperatures, grew from 78 days in 1952 to 95 days in 2011 [10].

The bottom line: latitude and tilt set the calendar of sunlight, and climate change leaves that calendar alone but adds energy and moisture on top of it. Summer now reaches later into the year, into September in London and New York and into October in Chennai, while the sunlight underneath still fades on schedule.

The other local factors that shape how hot we feel and how readily we sweat are wind, cloud, shade, and altitude. Wind has a particular cooling effect on the body. Much as the Earth is wrapped in its atmosphere, our skin is wrapped in a thin layer of still air that our body has warmed and moistened. Even a light breeze strips that layer away and replaces it with air that is cooler and drier, which speeds cooling in two ways: warmth passes more readily into the cooler air, and sweat evaporates more readily into the drier air. The first breath of wind helps the most; each added mile per hour helps a little less. In Marrakech, at 102 degrees Fahrenheit, the breeze brought air hotter than my skin, so it was delivering heat as well as taking it away. Whether it helped overall depended on whether the extra evaporation outweighed the heat it delivered.

Throughout Morocco, and again in Kensington, whenever my sweating got out of control, I reached for a small handheld fan. It was handiest where the air was still: in the air-conditioned van after a hot walk, or in a shop deep inside the medina, where the narrow lanes shut out the breeze. There it did exactly what a breeze does: it stripped away the warm, damp layer against my skin and replaced it with fresh air. In the 60-degree air of those London mornings, it could only help. Out in the open streets of Marrakech at 102 degrees, it had less to offer. Where the dry air was already taking my sweat as fast as I made it, a fan could add little more evaporation while blowing air hotter than my skin, and I was near the point where it does more harm than good. A 2021 modeling study proposed about 39 degrees Celsius, or 102 Fahrenheit, as an upper limit for fan use by healthy younger adults, with the exact limit depending on humidity [12].

Cloud droplets are much larger than the wavelength of visible light, so they scatter all its colors about equally, and that is why clouds look white. Air molecules are far smaller than a light wave and scatter blue far more than red, which is why the sky around the clouds is blue. By day, much of the light a cloud scatters goes back upward and never reaches the ground. On the overcast afternoon of my Thames walk, only 43 percent of the sunlight reached the ground, well below what a clear sky would have let through at that height of the Sun. By night, clouds work the other way. They absorb the infrared heat rising from the ground and radiate part of it back down, so the air beneath them cools slowly. Averaged over the whole planet and the year, the sunlight clouds turn away outweighs the heat they hold in [13].

Shade barely changes the temperature of the air around us, but it cuts the heat we absorb directly from sunlight hitting our body. Our skin and clothing absorb around 70 percent of the sunlight that reaches them, more for dark clothing and less for white. Seen from a Sun about 49 degrees above the horizon, as it was in Tangier that afternoon, a standing person presents to its beam an area of about a sixth of their skin [14].

In Tangier on September 25, we walked along the seafront in the shade, cooled by the breeze off the strait, and then crossed less than a mile of open, sunny street to the station. At 2:30 that afternoon, the direct beam of the Sun delivered about 658 watts to every square meter of flat ground, while scattered skylight added only 113 (ERA5 via Open-Meteo [15]). The shade had been blocking about 85 percent of the sunlight. Once we stepped out of it, by my calculation, my body began absorbing roughly 200 watts from the direct beam alone, plus a little from skylight and glare off the pavement. That is about two-thirds as much heat again as brisk walking makes, and on the walk to the station we had lost the sea breeze as well.

Altitude is the last of the local factors. On average, air cools about 3.6 degrees Fahrenheit for every 1,000 feet of climb [16], which is part of why Bangalore, about 3,000 feet up, escapes much of the oppressive heat that Chennai suffers at sea level. But height is only one lever. Marrakech sits about 1,500 feet up [2] and was still the hottest place on our trip, in part because its dry ground puts more of its sunlight into heating the air, and in part because it sits far inland from the cooling sea.

So, going back to my original question, “Was it me or the weather?” We now know after this comprehensive exploration of the science underneath, the answer has many layers. For starters, I produce heat within myself even as I walk; the sun and the hot air around me could add more; the Sun’s heat alone could differ based on where and when I am standing along with the angle of the Sun and how much the sky takes; my clothes, the breeze, and the water in the air decided how readily the heat escaped. What we felt may also vary based on body size. I had looked just at the thermometer and got confused.

By the end of our trip to Marrakech, my wife was done with the oppressive heat. She wants to plan better, making rational choices based on recent data so that we are comfortable on our vacations. It defeats the purpose of a vacation to suffer through it instead of relaxing and enjoying it. So, building on everything above, I have built a tool that estimates how comfortable a trip is likely to be, either by choosing the month and comparing places, or by choosing the place and finding its best months. I call it Physics Your Travel.

Physics Your Travel weighs the heat a walking body makes against the sun, the air temperature, the water in the air, the wind, and what we are wearing, for a brisk walk at the hottest time of day. It estimates two answers for each day, one in the shade and one in full sun, because stepping out of the shade can add nearly as much heat as the walk itself. It uses heat-balance constants from the international heat-stress standard (ISO 7933:2004 [17], since revised in 2023), starting from studies of how wet the skin can get before people feel uncomfortable [18]. The standard describes an average healthy adult; it does not predict any one person’s response, and neither does the planner’s year-round view built on it. For each place, half a month at a time, it estimates for a family which weeks might be comfortable, which may be sweaty, which could be oppressive, and which might be too hot for comfort at midday, in the shade and in the sun, along with what could be best to wear, how often it might rain, and whether the weeks could fall in storm season. The four labels come from one number, skin wettedness, the share of the skin that would have to be wet with sweat to shed the heat of the walk: comfortable below 0.36 [18], sweaty from 0.36 to 0.6, oppressive from 0.6 to 0.85, and too hot above 0.85, the most sweat an unadapted body can usefully put out by the standard [17]. The 0.6 line between sweaty and oppressive is my own working division, not a published threshold. These are measures of discomfort, not of danger; the planner does not judge whether a walk is safe.

The planner has two parts.

Its annual calendar view is built from the last ten years of records, 2016 to 2025, a compromise between keeping up with a warming climate and smoothing out odd years, and it needs rebuilding only once a year, when the year just ended joins the record and the oldest year drops out. It accounts for the recent weather baseline, what the last ten years say a given week is usually like. Climate science uses thirty-year averages; ten years follows a warming climate more closely but is noisier, since a single unusual year weighs more. It helps us decide where to go.

Then we have its forecast view for the next two weeks, which is downloaded anew every time we look up a place. It accounts for the weather, what the coming two weeks might look like. It helps us decide what to pack.

It has limits. Its calendar models one average adult walking briskly in light clothes. Its forecast can also take each of us into account: given a person’s height and weight, and a sweat test, it scales the average prediction to that person and estimates when they might start to sweat, how much, and what to wear. That scaling is my own heuristic, an exploratory calibration that has not been validated against measured sweating; the published equations behind it predict a group’s sweat rate, not any one person’s onset. When my wife and I take the test, we should be able to pack better for each of us, not for an average adult. A slower pace or a child may still feel it differently.

Its weather comes from airport stations, and a city center can run a few degrees warmer. It rates only the hottest part of the day, so mornings and evenings are kinder. And a forecast grows less reliable beyond about a week.

It was time to use the tool. Here is what it told us on a sample run. Our escapes from Phoenix to New York and Boston work best in late September, when about seven days in ten are comfortable for a midday walk in the shade in both cities; Boston’s best stretch begins earlier in the month, and New York’s lasts into early October. Those weeks still fall in the Atlantic storm season, though a tropical storm’s center has passed within 125 miles of either city in late September only about once in fifteen to twenty-five years [21]. Rome and Barcelona, which failed us in July, are comfortable on more than eight days in ten in October. Marrakech appears best from late November through March; in late September, when we went, almost no days were comfortable even in the shade. Chennai has no truly comfortable weeks for a midday walk at all; the least uncomfortable stretch is late December into early January, when the northeast monsoon rains have eased and a cyclone has passed within 125 miles only once in 45 years [21], but even then about eight days in ten are oppressive or worse in the shade. And London is comfortable on about eight days in ten right through the summer and into September.

Step into full sun, and every one of those best weeks shrinks. In New York and Boston, the comfortable days fall to two in ten or fewer; in Rome and Barcelona, to three in ten or fewer, and in Marrakech to about one in ten; and in London’s summer, to one or two in ten. In Chennai, full sun is too hot for comfort on nine days in ten or more, all year. The planner’s advice lies in the gap between its two answers: book the weeks that are comfortable in the shade, then keep to the shade at midday.

By this measure, the London days that set off this essay were comfortable ones in the shade, and our walks there were under cloud. That is the point: on those walks, it was mostly what I wore, not the weather.

This is a major improvement compared to how I planned my trips in the past. Science has given me a far more reliable instrument. Now, I wait for my wife’s verdict!

The planner my family uses to choose when and where to travel, Physics Your Travel, is now available as a paid tool for readers at physicsyourtravel.com. Paid subscribers to Tasting Status get 15 percent off.

The full version of this essay, with the formulas behind every number and every source, appears in The Quantastic Journal on Medium, at https://medium.com/the-quantastic-journal/the-physics-of-travel-1773d9450730

Sources

1. NOAA National Centers for Environmental Information. Global Historical Climatology Network daily (GHCN-Daily). https://www.ncei.noaa.gov/products/land-based-station/global-historical-climatology-network-daily

2. Iowa Environmental Mesonet, Iowa State University. Airport weather reports (METAR) and daily summaries: Marrakech Menara, Chennai Meenambakkam.

   https://mesonet.agron.iastate.edu

3. Met Office. Heathrow historic station data (monthly). https://www.metoffice.gov.uk/pub/data/weather/uk/climate/stationdata/heathrowdata.txt

4. NASA Science. Milankovitch (orbital) cycles and their role in Earth’s climate. https://science.nasa.gov/science-research/earth-science/climate-science/milankovitch-orbital-cycles-and-their-role-in-earths-climate/

5. NOAA Global Monitoring Laboratory. Trends in atmospheric carbon dioxide. https://gml.noaa.gov/ccgg/trends/

6. IPCC (2021). Summary for Policymakers. In Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press. https://www.ipcc.ch/report/ar6/wg1/chapter/summary-for-policymakers/

7. Myhre, G., Highwood, E. J., Shine, K. P., and Stordal, F. (1998). New estimates of radiative forcing due to well mixed greenhouse gases. Geophysical Research Letters, 25(14), 2715–2718. https://doi.org/10.1029/98GL01908

8. Forster, P. M., et al. (2026). Indicators of Global Climate Change 2025: annual update of key indicators of the state of the climate system and human influence. Earth System Science Data, 18(6), 3889–3933. https://doi.org/10.5194/essd-18-3889-2026

9. Held, I. M., and Soden, B. J. (2006). Robust responses of the hydrological cycle to global warming. Journal of Climate, 19(21), 5686–5699. https://doi.org/10.1175/JCLI3990.1

10. Wang, J., et al. (2021). Changing lengths of the four seasons by global warming. Geophysical Research Letters, 48(6), e2020GL091753. https://doi.org/10.1029/2020GL091753

11. ISO 7730:2005. Ergonomics of the thermal environment: analytical determination and interpretation of thermal comfort using calculation of the PMV and PPD indices and local thermal comfort criteria. International Organization for Standardization (revised 2025). https://www.iso.org/standard/39155.html

12. Morris, N. B., et al. (2021). Electric fan use for cooling during hot weather: a biophysical modelling study. The Lancet Planetary Health, 5(6), e368–e377. https://doi.org/10.1016/S2542-5196%2821%2900136-4

13. NOAA Geophysical Fluid Dynamics Laboratory. Cloud radiative effect. https://www.gfdl.noaa.gov/cloud-radiative-effect/

14. Arens, E., et al. (2015). Modeling the comfort effects of short-wave solar radiation indoors. Building and Environment, 88, 3–9. https://doi.org/10.1016/j.buildenv.2014.09.004

15. Open-Meteo. Historical weather API (ERA5 reanalysis).

    https://open-meteo.com

16. NASA Glenn Research Center. Earth atmosphere model. https://www.grc.nasa.gov/www/k-12/airplane/atmos.html

17. ISO 7933:2004. Ergonomics of the thermal environment: analytical determination and interpretation of heat stress using calculation of the predicted heat strain. International Organization for Standardization (revised 2023). https://www.iso.org/standard/37600.html

18. Fukazawa, T., and Havenith, G. (2009). Differences in comfort perception in relation to local and whole body skin wettedness. European Journal of Applied Physiology, 106(1), 15–24. https://doi.org/10.1007/s00421-009-0983-z

19. Glass, S., and Dwyer, G. B. (Eds.) (2007). ACSM’s Metabolic Calculations Handbook. American College of Sports Medicine; Lippincott Williams & Wilkins. https://books.google.com/books/about/ACSM_s_Metabolic_Calculations_Handbook.html?id=_wvBjsjkBU8C

20. Concept2. Pace-to-watts calculator. https://www.concept2.com/training/watts-calculator

21. NOAA National Centers for Environmental Information. International Best Track Archive for Climate Stewardship (IBTrACS), version 4, release 1. https://www.ncei.noaa.gov/products/international-best-track-archive

If you like this essay please restack it, so that others can enjoy it too.

It was September 12, 2026. I had just landed in London and went for a walk along the Thames in mid-afternoon after checking in to the hotel. It was special because I was catching up with long-lost friends from medical school. My wife is from the same class as me, so it was double the fun for us. We had been here before at this time of year and hence were sporting light jackets. As we walked down the river, I noticed I was starting to sweat. It was only a light sweat, but it surprised me, and the jacket went around my waist. Later that night, we were walking to my friend’s car, and again I felt the same light sweat. I was curious about why this should be happening at an ambient temperature in the high 60s and 70s Fahrenheit. Was it me or the weather?

We had then traveled to Morocco after a brief stint in Palma de Mallorca for a family wedding. Marrakech was wonderful but had oppressed us with some excruciating heat, making us question our choice of timing. Tangier at the tail end provided some respite. But then there was one other episode in Tangier, on September 25, when we walked from the beach, in the shade where it was cold, to the train station at around 2:30 p.m. During that walk in the open sun to the station, I had a sweating meltdown, with my body heating up perceptibly. I had to sit down in the station and cool myself with a hand fan. The others in my family did not have a similar experience. My son and my wife’s nephew had a milder version of it. Now I had so many questions. My naive mind was suffering from cognitive incongruence. Was it the weather, the climate, what I wore, or how much of it was me?

Then another strange thing happened on my return to London a couple of weeks later. It was the morning of September 27, and we were walking along Kensington Gardens. I was sporting a sweatshirt, under which I was wearing a regular T-shirt. After around 20 minutes of casual walking, I was telling my family that I might be at the sweating threshold. The others were wearing just T-shirts with open jackets. But my family ignored me as usual, and after another 10 minutes, I broke into some serious sweating. My family noticed this time and cut the walk short. When we got back to our hotel, right across from Cromwell Hospital, I had to cool myself with the hand fan and take off my sweatshirt. It took me probably 15 minutes before I could cool down to comfort. My family suggested that I was probably not in the best shape. But the very next morning, on a longer walk, we all went through the same episode together after about 45 minutes, dressed alike this time. The ambient temperature was in the 60s Fahrenheit on the first morning and only the high 50s to low 60s on the second. As a family, we had just arrived from Morocco, where we had experienced some oppressive heat. Now I had to dissect the science behind all this.

Before we dissect the science, let us understand our position in terms of sunlight hitting us, as that is the source of nearly all the heat that reaches us from outside. While we may think that we are on flat ground in London or Marrakech, from the sunlight’s perspective, we are on an incline. To put numbers on it, in London on September 12, the Sun at its highest point was about 43 degrees above the horizon, an incline of about 47 degrees. In Marrakech on September 21, it was highest at about 59 degrees, an incline of about 31 degrees. Even though we perceived the same flat ground we were standing on, we had a 16-degree difference in incline toward the Sun.

The Sun gives off about 3.83 × 10²⁶ watts of power [1]. By the time it reaches Earth, that power has spread over a sphere as wide as Earth’s orbit. The power hitting the top of Earth’s atmosphere facing the Sun is the Sun’s output divided by the area of that sphere, which works out to about 1,361 watts per square meter (W/m²), the value satellites measure [2]. It changes by about 3 percent either way because of Earth’s oval orbit.

Because Earth is a sphere, that beam of sunlight hits us at an angle in London and Marrakech. And because Earth’s axis is tilted by 23.44 degrees, that angle changes through the year. The share of the beam that lands on each square meter of ground depends on the sine of the Sun’s height above the horizon.

Let us sit a little with the sine. It is a number that falls from 1 when the Sun is straight overhead to 0 at the horizon. It is the fraction of the strength of the Sun’s beam that lands on each square meter of flat ground. The easy way to guess it is to look at the length of your shadow. When your shadow is as long as you are tall, the Sun is at a 45-degree angle, and the ground gets about 71 percent of the beam. If your shadow is longer than you, the Sun is lower and the ground gets less energy. If it is shorter, the Sun is higher and the ground gets more energy. At midday in London on September 12, my shadow was a little longer than I am tall. In Marrakech on September 21, it was just over half my height. Averaged across the whole planet, day and night, each square meter at the top of the atmosphere gets about a quarter of the solar constant, around 340 W/m², because Earth’s spherical surface is exactly four times the area of the disk it presents to the Sun.

Now comes the twist. Not all of that energy reaches the ground. The sunlight has to pass through our atmosphere first, and on the way it goes through what is called attenuation, a weakening of the beam as it travels through the air. First, the air molecules scatter light, blue more than red. That is why the sky is blue. Second, the particles in the air, mostly dust, smoke, sea salt, and pollution, scatter light as well, and together they make haze. We noticed airborne sand and dust doing this visibly in Marrakech. Then gases absorb light: ozone absorbs ultraviolet light, and water vapor and carbon dioxide absorb the infrared. Among the gases, water vapor takes the largest bite. Lastly, clouds reflect and scatter the light the most. Scattered light is not all lost; part of it still reaches the ground as diffuse light from the whole sky. For a single color of light, each layer of air removes a fixed fraction of what reaches it, so the losses compound like interest in reverse. Sunlight mixes many colors, and the easily removed ones go first, so the losses slow slightly as the path grows longer.

That is the theory of the atmosphere when the sky is clear and still. The real sky on those days had clouds and dust of its own. What matters is how much sunlight reached the ground where I was at those moments of our walks, set against what arrived at the top of the atmosphere above me.

Sunlight warms us in two ways. Our skin and clothes absorb it directly, and the ground absorbs it too. The ground takes in some of the light that reaches it and reflects the rest; the share it reflects is called the albedo. The ground then sheds what it absorbs in four ways: it radiates infrared heat, it warms the air touching it, it evaporates water from soil, leaves, and puddles, and it conducts heat down into the soil below. The air keeps warming for as long as the ground takes in more than it gives off, which lasts well past the moment the Sun is highest. So the hottest part of the day comes roughly two to four hours after solar noon, about 3 to 5 p.m. on summer clock time [4]. The Thames walk fell in that window; my Tangier walk came just before it. In Marrakech on September 21, it was 97 degrees Fahrenheit at 1 p.m. and 102 by 3 [4].

How the absorbed energy is split depends on how wet the ground is. Dry Moroccan ground has little water to evaporate, so far more of its energy went into heating the air than in damp London. Marrakech reached 102 degrees while its dew point, the temperature at which the air’s moisture would begin to condense and a measure of how much water vapor the air carries, sat near 54. That is where London’s was at midday on September 12, when its air was 70 degrees. The Moroccan air carried no less water vapor; it was simply much hotter. London’s ground was damper and spent much of its energy evaporating water, which heated the air less but made it wetter.

The inner city has more stone, brick, and asphalt, all of which absorb heat during the day and release it at night, and its buildings trap the night’s infrared and add waste heat of their own. That probably added to the warmth of our night walk. This is called the urban heat island effect, and in London it is mainly a nighttime effect. In summer monitoring, the center ran on average about 5 degrees Fahrenheit warmer than its rural surroundings at night, and on occasion up to 13 degrees [5]. There is another, slower effect. Around the equinoxes in late March and late September, the midday Sun stands at the same height over London, and an average September day over London gets only about 14 percent more sunlight than an average March day, measured above the atmosphere. Yet from 1991 to 2020 at Heathrow, September days averaged about 68 degrees Fahrenheit and March days about 53, and September nights were about as warm as March days [6]. What I felt in September was also heat that the ground and the sea had soaked up all summer, and in Britain, summer 2026 was provisionally the hottest in the Met Office record, which begins in 1884 [7]. They were still giving some of it back to the air.

Besides the heat itself, the other discomfort we feel is sweat on our body and clothes. Sweating is the mechanism evolution gave us to cool the body, but the cooling happens only when the sweat evaporates into the air. How fast it evaporates depends mainly on the gap between the water vapor at the wet skin and the water vapor in the air, and on how fast the air moves and what we wear, not on how hot the air feels.

This brings us to the concept of dew point. Dew point is the temperature to which the air would have to cool, without gaining or losing any water, before the water vapor it carries begins to condense into droplets, the way dew forms on leaves at dawn.

We often say that warm air “holds” more water. More precisely, the saturation vapor pressure, the most water vapor that can be present before condensation begins, rises with temperature. It nearly doubles with every 18 degrees Fahrenheit (10 degrees Celsius) of warming. The dew point tells us how much water vapor is in the air, whatever its temperature. Relative humidity, the number most weather apps show, tells us only how close the air is to its limit, and for sweat it misleads. At midday on September 12, London’s air was 70 degrees with a relative humidity of 57 percent. Marrakech on September 21 was 102 degrees with just 20 percent. Yet both had a dew point near 54 degrees, and so the same water vapor in the air. For skin at the same 93 degrees, the push driving sweat into the air was the same in both places, about 3.9 kilopascals. How much cooling I got also depended on wind, clothing, and how much sweat there was to evaporate.

At Pompeii in mid-July 2018, the temperature at the nearby Naples airport was about 90 degrees Fahrenheit, typical for a Naples summer. But the dew point reached 73 degrees [4], far higher than anything I would meet in London or Marrakech. On the open stone streets of the ruins, in full sun with almost no shade, the push driving sweat from skin at 93 degrees was only about two-thirds of what it had been in Marrakech’s 102 degrees, before counting wind and clothing. On that visit, one of our group was taken ill and had to be evacuated from the site.

We now move on to me, the human body. We have evolved to live in many habitats on this planet, and as a result we carry tools within us to deal with changes in our environment, including air temperature. The governing principle is homeostasis, the body’s way of keeping its internal state stable. For temperature, that means holding the core near 98.6 degrees Fahrenheit (37 degrees Celsius), give or take about a degree over the day. A small region at the base of the brain, the preoptic area of the hypothalamus, regulates it like a thermostat. It listens mostly to the core but also to the skin [8, 9]. When the core warms by even a few tenths of a degree, the brain turns on sweating and widens the blood vessels of the skin to carry heat outward, and warm skin makes these responses start sooner and run stronger.

The whole trip comes down to a balance sheet: heat builds up in the body when the heat it makes and gains outpaces the heat it loses. Moving air and radiation can add heat or take it away; evaporating sweat only takes it away. In Tangier, stepping into the sun added heat while losing the sea breeze made it harder to shed. In Marrakech, the air itself was hotter than my skin, so evaporation had to do almost all the work. In London, my sweatshirt held in the heat I was making as I walked. Different cities, the same balance.

The body’s metabolism makes heat. Just sitting, we make about 100 watts, like an old light bulb. Walking briskly, that rises to 300 to 400 watts [10]. On level ground, practically all of it ends up as heat; only when we climb does up to about a quarter go into lifting the body [11]. In the cool air of Kensington on September 27, my body was losing almost as much heat as it made. Under the sweatshirt, the small surplus added up slowly, and after about half an hour it crossed my sweating threshold.

In Marrakech, the air was 102 degrees Fahrenheit, hotter than my skin at about 93 degrees, so the air and the sun were both adding heat. Sweating was the only way out, and the dry air took the sweat as soon as it came. In London, the air was cooler than my skin and the sun was weaker, but my walking furnace made the same heat, and the sweatshirt tipped the balance. Under it, the sweat escaped only slowly, so it sat on my skin and soaked my T-shirt. The air trapped under the fabric filled with vapor from my own sweat and grew far more humid than the street outside; London’s open air that morning pulled on my sweat only about 7 percent less than Marrakech’s dry air had.

Sweat cools only when it evaporates, because turning liquid into vapor takes heat, and it takes that heat from the skin. When the air is as hot as the skin, shedding 300 watts of walking heat takes about half a liter of sweat an hour; in cooler air it takes less, because the air carries some of the heat away directly. Sweat that drips off cools almost nothing, and sweat that soaks into clothing cools less than sweat on the skin, because it evaporates farther from the skin. So the body makes more to compensate.

Why should escaping water take heat with it? The molecules in a film of sweat are always jostling, at a spread of speeds. Only the fastest have enough energy to break free of their neighbors and escape into the air. Each one that leaves takes its energy with it, so the molecules left behind are slower on average, and slower means cooler. For every gram of sweat that evaporates at skin temperature, about 2,420 joules leave with it [12]. A drop that rolls off takes almost nothing, because it never broke free. The clay water pot kept in Indian homes through the summer, the paanai in Tamil, works the same way, and I keep one at home in Phoenix. Water seeps through the unglazed clay and evaporates from its outer surface, cooling the clay and the water inside. The pot and the skin are the same machine.

On that half-hour walk on September 27, the rest of my family did not sweat nearly as much, because they had not dressed as warmly as I had. They had also just come from Morocco, so whatever heat training our bodies carried back, we shared. The difference was what we wore. My clothes held in part of my body heat, so my core temperature climbed faster and crossed the sweating threshold sooner. And once the sweat came, the same layers slowed its escape to the open air, so it cooled me less than theirs cooled them.

The next morning, September 28, we were all dressed alike, and the air was cooler and drier, near 60 degrees. None of us sweated for the first three-quarters of an hour. It came on the final stretch, after a short stop in a wine shop, and then it came for all of us together. Even then, my son and I sweated more than the others. Clothing and weather are not the whole story. Bodies differ too. My wife’s nephew, bigger and fitter than the rest of us, had also sweated more than most of us in Tangier. A larger body makes more heat for the same walk, and sweats more to shed it [13].

Besides constitutional differences, there is yet another human variable that shapes how we respond to the temperature around us: acclimatization. Our bodies adapt to where we live long-term, and also to where we have recently spent time. We had just spent close to ten days in Mallorca and Morocco. And my family and I have lived in the Phoenix area for the last eighteen years.

Acclimatization to hotter environments makes a person start sweating at a lower core temperature. The sweat comes sooner and in larger amounts, and it carries less salt. Blood volume expands, and the heart works less hard in the heat. All of it serves one end: to start cooling the body sooner and keep the core from climbing. The adaptation builds over one to two weeks of daily heat exposure [14] and starts to fade as soon as the exposure stops [15]. Our six days in Morocco’s heat may have lowered our sweating threshold a little by the time we returned to London on September 27, but on that walk the sweatshirt alone is enough to explain my sweating. Our long years in Phoenix matter only as far as we spend time in its heat, and mostly we live in our cold cocoons at home and at the office.

Coming next, in Part 2: how the climate has shifted under my travels, then wind, shade and altitude, and the planner I built from all of it.

The full version of this essay, with the formulas behind every number and every source, appears in The Quantastic Journal on Medium: https://medium.com/the-quantastic-journal/the-physics-of-travel-1773d9450730?sharedUserId=shanmukund

Sources

1. Prša, A., et al. (2016). Nominal values for selected solar and planetary quantities: IAU 2015 Resolution B3. The Astronomical Journal, 152(2), 41. https://doi.org/10.3847/0004-6256/152/2/41

2. Kopp, G., and Lean, J. L. (2011). A new, lower value of total solar irradiance: evidence and climate significance. Geophysical Research Letters, 38(1), L01706. https://doi.org/10.1029/2010GL045777

3. Meinel, A. B., and Meinel, M. P. (1976). Applied Solar Energy: An Introduction. Addison-Wesley. https://archive.org/details/appliedsolarener0000mein

4. Iowa Environmental Mesonet, Iowa State University. Airport weather reports (METAR) and daily summaries: London Heathrow, Marrakech Menara, Naples Capodichino. https://mesonet.agron.iastate.edu

5. Watkins, R., Palmer, J., Kolokotroni, M., and Littlefair, P. (2002). The London heat island: results from summertime monitoring. Building Services Engineering Research and Technology, 23(2), 97–106. https://doi.org/10.1191/0143624402bt031oa

6. Met Office. Heathrow historic station data (monthly). https://www.metoffice.gov.uk/pub/data/weather/uk/climate/stationdata/heathrowdata.txt

7. Met Office (2026). Summer 2026 provisionally hottest on record for the UK. https://www.metoffice.gov.uk/about-us/news-and-media/media-centre/weather-and-climate-news/2026/summer-2026-provisionally-hottest-on-record-for-the-uk

8. Egan, G. F., et al. (2005). Cortical, thalamic, and hypothalamic responses to cooling and warming the skin in awake humans: a positron-emission tomography study. Proceedings of the National Academy of Sciences, 102(14), 5262–5267. https://doi.org/10.1073/pnas.0409753102

9. Ravanelli, N., Gendron, P., and Gagnon, D. (2021). Revisiting the evaluation of central versus peripheral thermoregulatory control in humans. American Journal of Physiology: Regulatory, Integrative and Comparative Physiology, 321(2), R91–R99. https://doi.org/10.1152/ajpregu.00321.2020

10. Herrmann, S. D., et al. (2024). 2024 Adult Compendium of Physical Activities: a third update of the energy costs of human activities. Journal of Sport and Health Science, 13(1), 6–12. https://doi.org/10.1016/j.jshs.2023.10.010

11. Minetti, A. E., et al. (2002). Energy cost of walking and running at extreme uphill and downhill slopes. Journal of Applied Physiology, 93(3), 1039–1046. https://doi.org/10.1152/japplphysiol.01177.2001

12. Lemmon, E. W., McLinden, M. O., and Friend, D. G. Thermophysical properties of fluid systems (saturated water). NIST Chemistry WebBook, NIST Standard Reference Database 69. https://webbook.nist.gov/chemistry/fluid/

13. Cramer, M. N., and Jay, O. (2015). Explained variance in the thermoregulatory responses to exercise: the independent roles of biophysical and fitness/fatness-related factors. Journal of Applied Physiology, 119(9), 982–989. https://doi.org/10.1152/japplphysiol.00281.2015

14. Sawka, M. N., Périard, J. D., and Racinais, S. (2016). Heat acclimatization to improve athletic performance in warm-hot environments. Sports Science Exchange, 153. Gatorade Sports Science Institute. https://gssiweb.org/sports-science-exchange/article/sse-153-heat-acclimatization-to-improve-athletic-performance-in-warm-hot-environments

15. Daanen, H. A. M., Racinais, S., and Périard, J. D. (2018). Heat acclimation decay and re-induction: a systematic review and meta-analysis. Sports Medicine, 48(2), 409–430. https://doi.org/10.1007/s40279-017-0808-x

Thanks for reading Subliminal! Subscribe for free to receive new posts and support my work.

Subscribe now

If you enjoyed the essay, please click the heart and restack it so others can enjoy it too.

This essay first appeared in The Subliminal, my science publication. I am sharing it here because it will help anyone who travels.

It was an unusually sweltering late September afternoon in Marrakech, with the temperature reading at 104 °F. I was headed to Mechoui Alley deep inside the maze of the medina. Of course, I needed a guide as I would have been lost after a few short minutes. My guide was a wonderful person who besides having a beautiful soul had done the homework of courting the chef I was going to see, days before I landed. Now, before we get lost in the maze of the medina, you might be asking yourself: what is méchoui?

Méchoui comes from the Arabic word mashwī, which means roasted. A 13th-century cookbook from Muslim Spain and North Africa already gives recipes for roasting a whole lamb in the oven [1]. The oven made out of clay originally comes from Mesopotamia. It is called the tannūr. This word goes back to the Akkadian word tinūru, which means oven. The word tandoor comes from its Persian cousin, tanūr. The roasting of whole animals such as lamb in these ovens goes as far back as the 10th century AD in Baghdad. Once the fire dies down to embers in the oven, a lamb is hung over a pot in the middle of the tannūr, basted with salt, and the oven is sealed with clay. The 10th-century cookbook Kitāb al-Ṭabīkh, which means the book of cooked food, written by al-Warrāq, gives the recipe for a whole lamb hung in the middle of a tannūr over a pot, basted with salt, and the oven sealed with clay [2].

From al-Warrāq’s Kitāb al-Ṭabīkh, chapter 91, “On tannūr dishes”

فإذا سكن التنور أخذ الجدي أو الحمل فينظف وتجعل القدر وسط التنور ويعلق عليها ويسقى بملح ويطين.

When the tannūr has died down, take the kid or the lamb and clean it. Set the pot in the middle of the tannūr and hang it over the pot. Baste it with salt and seal it with clay.

Later, the same 13th-century cookbook gives a similar recipe of a lamb breast hung in the tannūr with a vessel underneath “into which whatever melts from the breast drips” [1].

From the 13th-century Andalusian cookbook, “Another kind of lamb breast”

نوع آخر من الجنب. يؤخذ جنب خروف سمين … يخاط الجنب بمصران نظيف ويعلق في التنور ويجعل تحته آنية يقطر فيها ما يتحلل من الجنب فإذا كمل أخرج.

Another kind of lamb breast. Take the breast of a fat lamb. … Sew the breast shut with a clean gut and hang it in the tannūr. Set a vessel beneath it into which whatever melts from the breast drips. When it is done, take it out.

In the ancient past several cultures were cooking meat in a pit under the ground. The Mongols were pit roasting whole animals, as documented in a Mongol court cookbook presented at court in 1330. A whole sheep with fleece on was laid on an iron grate in a pit about a meter deep lined with red-hot stones, covered with willow branches and sealed with earth [3]. Their everyday fuel was dried dung from their herds [4]. But dung needs oxygen to burn and would not be of much help in closed pits. So, they used a two-part strategy. First, they would heat stones with burning dung to make them red hot. Then they would line the underground pits with these stones and seal the opening for the meat to roast in that heat [5]. Even though dung was freely available from their herds grazing on the pastures, it was still a scarce commodity as they had to store it to heat their homes through harsh winters [6]. So roasting whole animals was reserved for feasts, while everyday meat was cut up and boiled in a cauldron [4][7].

Coming closer to Marrakech, the Bedouin of Arabia also cooked meat in pits under the ground. They too had fuel scarcity, living in an almost treeless landscape of their own. Their smaller herds likely gave less dung. They cooked chopped or jointed meat in underground pits with hot stones, sometimes packed inside an animal’s stomach, then covered it with hot ashes [8][9].

Back in the meandering alleys of the medina, my guide talked fast as he explained that the shop we were going to was supposed to open in the morning but they decided to open late that day. We passed through several bustling alleys busy with tourists bargaining with the shop vendors who can be pretty aggressive. While dodging all that, we passed by a restaurant that caught my eye. It was Chez Belhaj, under a sign that read Tanjia Mrrakchia, Marrakech tangia. There were a bunch of dishes called tangia, which are essentially small clay urns of meat cooked in hot embers, sealed with paper.

Paper-sealed tangia urns at Chez Belhaj, Mechoui Alley, Marrakech. On the menu: head meat, trotters, meat and chicken. Photo: the author.

Going past that restaurant, the medina never ceased to intrigue me. I saw a bunch of men haggling with a man on his bicycle with a bunch of chickens stacked at the back. They were arguing about price. When I asked my guide, he said this gentleman brings in farm-raised chickens, called beldi, meaning “of the country,” which fetch a premium here over the broilers sold in the markets. He cycles from villages outside the city. The haggling was over the premium he charged for them. The sale failed and he wheeled his bicycle away while I caught him on my phone.

After the failed sale: live beldi chickens tied in the crate as the seller wheels away through the medina, Marrakech. Photo: the author.

Passing that, we came across another interesting vendor who was stacking up bread dotted with semolina. He had a public bakery, a farran, the neighborhood’s communal oven, where families from the community bring their dough, often marked so they can tell their loaves apart, and he bakes it for them for a small fee. The leftover loaves were on sale as well. This is the town side of the story: the settled people’s oven, baking the community’s bread, just as the tannūr did long before the nomads used it to roast whole animals.

This is the town side of the story: the settled people’s oven, baking the community’s bread, just as the tannūr did long before the nomads used it to roast whole animals.

Loaves from a communal oven, dusted with semolina, for sale from a cart in the medina, Marrakech. Photo: the author.

Finally, we took the turn towards our target, passing by what I would call the most spectacular showcase of olives I have seen. You could see the faint mist covering my phone camera from the oppressive humidity I was experiencing then.

Jnane Zaitoun, “olive garden,” shop No. 10: mounds of red, green and black olives below jars of preserved lemons and peppers, in the medina, Marrakech. Photo: the author.

Our target was the restaurant Chez Lamine Hadj Mustapha, run by Al-Amīn al-Hajj Mustafa an-Nakīr, a legendary figure who, by his own account, comes from a family of cooks. His restaurant specializes in méchoui. It is a humble place flocked by people, and several locals were lined up to collect the meat to take it home.

The front of Chez Lamine Hadj Mustapha, under a wall of visitor photos, one labeled Gordon Ramsay. Photo: the author.

Méchoui carved to order and weighed on a brass scale, with round bread and a bowl of cumin and salt. Photo: the author.

Tangia urns, sealed with paper and foil, outside Chez Lamine in Mechoui Alley. Photo: the author.

Many tourists from far and wide were sitting in the restaurant to eat their bite, and a large bite at that! I was taken to the chef directly as my guide had done the handshake earlier and indicated that I was a food aficionado. He greeted me warmly and wanted to check out my Substack profile. Satisfied, he took me on a tour around his restaurant and showed me the pit where they cook the lamb.

The pit’s clay mouth set into the tiled floor: a lamb hangs over pans that catch the dripping fat. Photo: the author.

Then he very generously gave me a very good piece, hand-selected by him, served the classic way with bread, and cumin and salt on the side. He then seated me at the restaurant and departed after a quick photograph.

With Al-Amīn al-Hajj Mustafa an-Nakīr inside his restaurant, Chez Lamine Hadj Mustapha, Marrakech. Photo: the author.

We will talk about him later. First, the meat.

What landed on my plate was a rib section that the chef picked out himself, sending back every other piece his men brought until he had the one he wanted. He handed me the plate with his right hand, the hand of honor in Moroccan hospitality. It was absolutely delicious and frankly too much for one person. But I devoured it anyway. The accompaniments were simple: cumin powder mixed with salt to sprinkle on, and bread to eat it with, used as the spoon, all on parchment paper, which was perfect for the fat that was spilling ever so slightly from the meat. The interesting thing was that the skin was roasted crisp without browning too much, which was a delight by itself.

The chef’s pick, with round bread and cumin mixed with salt, on paper. Photo: the author.

Inside the cut: crisp skin over soft fat and meat that pulls apart. Photo: the author.

Let us talk about how the meat got cooked into this deliciousness. The underground oven you saw earlier had hardwood burnt all the way to embers in it. The heat that emanates soaks into the clay walls and the packed earth around them. Clay and earth store about 0.9 kilojoules per kilogram for every degree they warm. In a typical pit in Mechoui Alley that measures approximately 2 meters deep and 1.4 meters across, a morning fire heats only the inner skin of the wall, about 10 centimeters deep, but that is still about 1,700 to 2,100 kg of clay. Warmed an average of 250 °C above the air, it holds about 380,000 to 470,000 kilojoules of energy, enough, on paper, to cook about 30 to 40 lambs, though the pit only holds a few. Imagine that. That stored heat cooks the lamb for hours after the flame is long gone. Being underground adds another bonus factor. Dry earth is a poor conductor of heat and sends most of the heat inward instead of losing it to the air, and sealing the mouth keeps the hot air from rising out. That heat gets to the meat in two ways. Radiation from the hot walls, which grows with the fourth power of their temperature (the Stefan-Boltzmann law), reaches the meat directly through the air. The hot trapped air itself heats it by convection. The lamb is hung in the center to be cooked evenly on all sides without having to turn the meat. As I said, the heat radiation follows the fourth power of the wall temperature. The walls cool slowly, but every degree they lose cuts the heat sharply, so the meat cooks ever more gently. It is a falling-temperature oven, not a steady one. That is physics, but nobody has measured the exact drop rate.

That stored heat cooks the lamb for hours after the flame is long gone.

Only the inner skin of the clay wall stores the fire’s heat; its radiation grows with the fourth power of wall temperature (the Stefan-Boltzmann law). Illustration: the author with Chat GPT.

The next force in this cooking is humidity. The interesting thing is the moisture in the meat itself provides it, as about 63 percent of lamb meat is water [10]. In a sealed chamber, the water vapor has nowhere to escape and hence the air inside the oven turns humid. It is just like how the medina itself is so much more humid compared to the dry heat outside in the city. Humid air cooks faster than dry air. When the water vapor condenses on the cooler meat, it releases about 2,260 kilojoules of energy per kilogram of water. The surface stays moist and cooks uniformly. But eventually, the pit is opened at the end and left for a while, letting the remaining moisture escape and the skin crisp, cooked from underneath. That crisp skin is the cooks’ sign that the meat is ready.

In the sealed pit, steam condensing back on the meat releases about 2,260 kJ per kilogram of water. Illustration: the author with Chat GPT.

The coating is the last trick to this method. The smen, which is aged butter, conducts heat to the skin better than air and fries the surface. The coating also slows moisture loss, which helps keep the meat moist. The salt pulls out the water from the skin by osmosis, helping it dry and crisp. It also helps loosen the surface proteins. Cumin’s aroma compounds dissolve in the smen, carrying the flavor into the skin, which is why I tasted it distinctly.

Smen, salt and cumin, the coating the chef confirmed, each do a different job on the skin. Illustration: the author with Chat GPT.

Now for my actual piece that I consumed. It was rib meat, which has a generous amount of fat and a thin stretched skin on top, which is perfect for crisping. The fat beneath the skin melts and bastes the meat, which makes it that much more succulent and delicious. The rib’s connective tissue, called collagen, starts to break down at about 53 to 63 °C and, over hours of slow heat, turns to gelatin, which holds water [11]. That is why the meat came apart so easily and stayed juicy.

Skin, fat, collagen and bone each change under slow heat; collagen starts breaking down at about 53 to 63 °C [11]. Illustration: the author with Chat GPT.

Finally, as I promised, let us talk about the man himself. Al-Amīn al-Hajj Mustafa an-Nakīr is something of a legend in Mechoui Alley. The locals told me he once cooked for the previous king, and a 2018 Moroccan news headline calls him “Hassan II’s cook” [12]. He says he comes from a family of cooks, and his own accounts of when he served the palace and for how long vary from one interview to the next, as stories that old often do. What I can vouch for is the man I met. He was warm and curious, checked out my Substack before he trusted me with his kitchen, walked me to the pit himself, and sent back every piece of lamb his men brought until he found the one he wanted me to have. Whatever the exact years at the palace, he still treated me with honor just like a special guest with a special offering, and for that I am grateful.

Whatever the exact years at the palace, he still treated me with honor just like a special guest with a special offering, and for that I am grateful.

The chef has been quoted as saying that méchoui comes from the Sahara, and that might be what the voice of tradition says. But we have evidence that points the other way, to the tannūr of the towns. It could very well be that most Moroccan towns have given up the pit, while the countryside kept it, so the method now looks as if it came from the desert, the reverse of how it began. In Marrakech, Mechoui Alley was the only place I could find it; anywhere else meant an hour or more into the countryside, as my guide and the locals confirmed. That is one more reason to be grateful to the chef for keeping this tradition alive and accessible to people like me.

But there was this one question nagging at me. I had to ask the local while I waited for my guide. “Why, if most of the dishes I encounter in Morocco, including méchoui, are found in an Arabic text written more than a thousand years ago, do I not see them anywhere else in the Arab world?” I asked. “Aha, that is a good question!” he said. He then explained that the Arab influence in cuisine marched through North Africa, including Morocco, all the way up to al-Andalus. But it got lost in the Arab heartland itself because the Ottoman Empire replaced it with its own. “This is why you need to come to Morocco,” he said with a smile and a wink, as my guide approached me to take me back to the cool oasis of my riad. The local gentleman was half right: Morocco never became Ottoman territory [13], and its kitchen kept more of the old Arab table, though pit roasting still survives in Yemen’s mandi [14] and the Bedouin zarb [15].

In Mechoui Alley, méchoui is something older than the word. It is the clay oven of the first towns of Mesopotamia, carried west by the Arabs across North Africa and into Spain, sunk into the ground, and filled with a whole lamb the way a feast demands. It is the patience of a single fire that cooks a single batch, the science of stored heat, steam and slowly melting fat, and the hands of a man who still chooses the best piece for his guest. I came looking for a dish and found a thousand-year-old conversation between the town and the desert, still going on in a pit under a tiled floor in Marrakech.

Sources

1. Anonymous, An Anonymous Andalusian Cookbook of the 13th Century, trans. Charles Perry. Recipes “Roast Lamb,” “Lamb Roast with Its Skin” and “Another Kind of Lamb Breast.” Arabic original: Paris, Bibliothèque nationale de France, Arabe 7009.

2. Ibn Sayyār al-Warrāq, Kitāb al-Ṭabīkh, 10th century, Baghdad, chapter 91. Ed. Kaj Öhrnberg and Sahban Mroueh, Studia Orientalia 60 (Helsinki: Finnish Oriental Society, 1987).

3. Hu Sihui, Yinshan zhengyao (Proper and Essential Things for the Emperor’s Food and Drink), presented at the Mongol court of China, 1330. Recipe “willow-steamed lamb.”

4. William of Rubruck, The Journey of William of Rubruck to the Eastern Parts of the World, 1253-55, trans. W. W. Rockhill, 1900.

5. S. L. Black and A. V. Thoms, “Hunter-Gatherer Earth Ovens in the Archaeological Record: Fundamental Concepts,” American Antiquity 79, no. 2 (2014): 204 to 226. https://doi.org/10.7183/0002-7316.79.2.204

6. D. Lkhagvadorj, M. Hauck, C. Dulamsuren and J. Tsogtbaatar, “Pastoral Nomadism in the Forest-Steppe of the Mongolian Altai under a Changing Economy and a Warming Climate,” Journal of Arid Environments 88 (2013): 82 to 89. https://doi.org/10.1016/j.jaridenv.2012.07.019

7. P. D. Buell, M. Oskenbay, M. de Pablo Moya and E. N. Anderson, Crossroads of Cuisine: The Eurasian Heartland, the Silk Roads and Food (Leiden: Brill, 2020). https://doi.org/10.1163/9789004432109

8. al-Azharī (died 980), Tahdhīb al-Lugha, quoted in Ibn Manẓūr, Lisān al-ʿArab, entry كرش.

9. Ibn Sīda (died 1066), al-Muḥkam, entry حنذ.

10. U.S. Department of Agriculture, FoodData Central, “Lamb, composite of trimmed retail cuts, separable lean and fat, trimmed to 1/8” fat, choice, raw,” FDC ID 172545.

11. Eva Tornberg, “Effects of Heat on Meat Proteins: Implications on Structure and Quality of Meat Products,” Meat Science 70, no. 3 (2005): 493 to 508. https://doi.org/10.1016/j.meatsci.2004.11.021

12. Hespress, “طباخ الحسن الثاني” (Hassan II’s cook), July 28, 2018. https://www.hespress.com/طباخ-الحسن-الثاني-442022.html

13. Jamil M. Abun-Nasr, A History of the Maghrib in the Islamic Period (Cambridge: Cambridge University Press, 1987).

14. Embassy of the Republic of Yemen in Poland, “Must try.” https://embassy-of-yemen.pl/en/home/visit-yemen/tourism-in-yemen/must-try/

15. Yasmin Noone, “Underground Bedouin BBQ is addictively delicious,” SBS Food, February 9, 2017. https://www.sbs.com.au/food/article/underground-bedouin-bbq-is-addictively-delicious/dld4qr6yo

If this took you somewhere, restack it so others can walk the alley too.

We settled down in the town of Bruges. It might be a city by pure demographic definition but you do not perceive it. Its medieval Flemish architecture is charming. It gives you a vibe of an idyllic village with facades adorned by flower pots on window sills and manicured lawns with eclectic gnomes in the homes that line the picturesque canals. This was instantly therapeutic to our minds escaping the chaotic blend of old world displaying colonial wealth and the new world of streets with expressive graffiti, disorderly traffic and strong multicultural mosaic in Brussels.

I have personally been here before mostly because of my love for Flemish beer and Belgian chocolates. I had not treated the local cuisine seriously before. This time was different. My wife and my son were traveling with me. My wife is a vegetarian and my son loves starchy food. Croquettes and steak were not going to cut it. I was motivated to explore food options and I found a couple of Flemish stews to be surprising choices to satisfy my family’s food diversity.

We were staying at Hotel De Orangerie, a boutique hotel at canalside. It was a converted convent of the 15th century, a perfect representation of this city’s past in how the decor and the rooms were designed. We checked into our room for a brief break and started our excursion in the city exploring flights of beer at the Beer Wall and immersed ourselves in chocolate tasting later.

Then as promised, we were headed to the restaurant of my choice, Cocotte. It was situated off the tourist beat. I had made this reservation several weeks in advance. We were greeted by a wiry gentleman. We were early for dinner. “Please come in. Pick wherever you want to sit.” He pointed to the empty restaurant with a wry smile. We smiled as we sat down in our chosen chairs. What captured our attention was the wallpaper and what appeared to be an old cooking stove with vintage colors, all of which transported us back to the early 1900s. More importantly, before any of this, the wonderful aromas of the stews greeted us.

We all peered down at the menu, zeroing in to confirm they had what we had decided earlier from the website. My wife ordered the ratatouille, a vegetable stew, my son the kip cocotte, chicken with meatballs, and I the konijn, rabbit braised in beer with prunes and bacon.

He told us that all of that would be ready in about 9 to 10 minutes. We were taken aback a bit as we had imagined that the raw ingredients would take a while to stew in one of those iron pots he had on display on the old stove top.

He then explained to us that he had a head start every day of having prepared the stews a day earlier that would need a reheat in those iron pots for 9 to 10 minutes to be served just like grandmother would do.

Just like he promised, our dishes arrived in 10 minutes. They looked inviting and the aroma of the stews worked up our appetite. Both smell and sight of food trigger salivation through a reflex. Our body preps our digestion before we consume just like a kitchen preps the ingredients before cooking. The aroma comes from compounds that the heat drives out of the dish. Most of them dissolve in fat and barely at all in water, so a stew built on bacon fat or olive oil holds far more of them than a lean broth ever could, and keeps releasing them instead of giving everything up at once. The lid holds them in the space above the food rather than letting them escape, so they are still there when the pot reaches the table. It is the appetizer that we don’t consume.

It is the appetizer that we don’t consume.

My wife indulged in her dish with exuberance as a wholesome vegetarian dish like this is not that common here. Her face lit up as she started to consume the stew. The tomatoes had broken down into the sauce, which coated the wedges of green and yellow pepper, aubergine and courgette, along with the olive oil. The feta cheese cubes were littered like white soft boulders on top. The pectin in vegetables starts softening above 80 to 90 degrees Celsius. But the acid in the tomato slows pectin breakdown, which is what helps to hold the shape of the vegetables. Glutamate in the tomato and onion adds the umami flavor. The acid has already set the proteins in feta, so they soften without melting. The salt suppresses any sourness and makes sweetness read stronger. My wife noted that she tasted the sweetness from peppers, warmth from the black pepper, and the feta balanced the sweetness and sourness of the tomato. She felt the umami flavors from onion, tomato bathed in olive oil.

A Provençal vegetable stew cooked by the same method, in the same pot, as the beef and rabbit. Delivered.

My son’s dish came with a different appearance altogether. He was famished and was ready to pounce on that stew. A thick creamy velvety beige chicken stew with raw red tomato halves, charred green asparagus, spring onion and croutons scattered across lay in the pot in front of him. Tucked in between all of that were meatballs coated with that sauce. My son’s pace of eating started to quicken, which is his way of saying that the food was delicious. I poked a little into his dish to get a sample. I was immediately immersed in its deep umami flavors emanating from chicken and mushroom in the cream sauce along with the sensory pleasure of the crunchy croutons mixed with the softened ones layered on top. The sauce had a roux base. As the heat increases the starch granules absorb water and swell. The fat then coats each granule keeping them separate. The thickness of the sauce is the swollen starch and the velvetiness is the fat coating the granules. Both chicken and mushroom are high in glutamate. The mushrooms release a lot more of them as liquid into the sauce as they are mostly water. The additional nucleotides in the meat have a multiplier effect with glutamate, which creates a lot more savoriness or umami when they are paired. The myosin in the meatballs is dissolved by the salt. The milk-soaked bread sitting between the meat strands prevents tightening keeping them soft. They also contribute fat and juices to the sauce. The croutons undergo the Maillard reaction and act as incidental soakers of sauce around them. The raw tomato, spring onion and parsley go in as garnish at the end. The potato was the glistening side dish delivered after roasting in the oven.

Now we land on my dish, the rabbit stew. It was a stew with a sauce of dark reddish brown color, similar to a Flemish beer, with the paler rabbit meat sitting in it and prunes scattered around the pot. You could see quite a few bacon lardons in the sauce. The dish was garnished with parsley on top.

Rabbit is one of the leanest meats, with almost no fat marbled inside the muscle. Its proteins set in two stages. Myosin sets at around 50 degrees Celsius, which does no harm. Actin sets between 66 and 73 degrees Celsius, and that is the stage that squeezes the water out and leaves the meat dry and stringy. With no fat inside the muscle to mask the loss, rabbit punishes overcooking faster than beef or pork. This is what the bacon lardons are for. They render their fat into the pot, and that fat coats the muscle fibers and carries flavors that water cannot. The technique is called barding, and cooks have used it on lean game for centuries.

The cooking starts with the lardons alone in the pot, rendering their fat, which becomes the medium everything else browns in. They come out, and the rabbit pieces go in to brown in that fat, then come out too. The onions go in after them and sweat in the same fat until they soften. Then the beer, which lifts the fond, the browned residue stuck to the bottom, back into the liquid and turns it into the base of the sauce. The rabbit and the lardons return, the herbs with them, and the pot goes onto low heat for the long slow cook.

The onion goes in only after the rabbit comes out, because onion is about 88 percent water. Put in alongside the meat, that water floods the pot and drops the temperature below what browning needs, so the meat steams instead. On its own it wants exactly that wetter, cooler pot. Its water dissolves the fond and its sugars break down slowly into sweetness. The prunes come later, and later is enough for them. They are already dried and collapsed before they meet the pot, so they soften quickly in hot liquid, and their sugars dissolve into the sauce fast. What they do not have time for is pectin, which needs longer heat to thicken anything. I do not like fruit in a meat dish. I ate every one of these.

I do not like fruit in a meat dish. I ate every one of these.

Let us talk about why beer, and why this beer. When a sauce reduces, water leaves and everything else concentrates in proportion. Halve the liquid and you double the sweetness and the bitterness alike, because reduction cannot choose what to keep. The beer chosen here was Bernardus Abt 12, a quadrupel, a strong dark Belgian ale with very low hop bitterness. Reducing it concentrates malt sweetness and dark fruit, because that is what is in there to concentrate. A hoppy beer in the same pot would have turned harsh.

Belgian abbey yeasts also produce esters during fermentation, aroma compounds that smell of dried fruit and plum. These complement the aroma coming from the prunes, so the two arrive as one family of flavors by two different routes.

The alcohol has its own part. Contrary to popular opinion, not all of it burns off in cooking. Around 25 percent survives an hour of simmering, 10 percent two hours, and 5 percent two and a half, and a lid holds back more as the vapor condenses and falls back in. What stays dissolves aroma compounds that water cannot, letting us perceive flavors that would otherwise be lost.

The chef here uses only the legs. Legs are full of collagen, the connective tissue that makes meat tough, and above 60 to 70 degrees Celsius it slowly converts into gelatin over hours. That is why he braises a day ahead. It is also why reheating does not toughen the meat, since the tough part has already become the sauce. That gelatin is what gives the sauce its body and its gloss, with no flour anywhere in it.

The delay is a gain rather than a cost

The word cocotte means a pot for slow cooking, and it also means a hen. The two meanings come from separate origins and drifted into each other over centuries, so the pun in the name is one the language made rather than one anybody chose. In the restaurant the pot is plainly what is meant. The sign on the wall, Geluk zit in een klein potje, happiness sits in a little pot, is the deliberate one. It plays on the Dutch proverbs about small corners and about the best ointment coming in the smallest pots, and geluk carries both luck and happiness at once, which no single English word will hold. The stove on display is a Leuvense stoof, a nineteenth century Leuven stove, fitted with a modern induction top. The enamelled cast iron pots standing on it are there to be looked at, not cooked in, and that style of pot is itself a Belgian invention, made by two Belgians in northern France in 1925 and read ever since as French. The recipes are the ones Willem ate at his grandmother’s table. She cooked them in a stoofpan, a plain stewing pan, not in these. He spent years as an independent professional in information technology before this, and he closed for a while to care for his mother. What he is doing now is keeping Flemish home cooking alive by cooking it.

As we ate, it was hard not to look at the room. Ornate wallpaper, and on it an oval photograph of a woman and her husband. I had taken both for period styling, a restaurant dressing itself in its era. Willem told us the wallpaper had been on the ceiling of his grandmother’s house, and that the photograph was of his grandparents. We instantly realized that this man is recreating these dishes from a deep emotional memory embedded in him, and we get to participate in that.

One man can send out three different stews to three different tables, alone, in ten minutes.

Slow food has its challenges. The customer arrives before the dish is ready, which is why restaurants are built around several cooks working at once. A braise made the day before reverses that twice over. It is finished long before anyone walks in. And the dish gets better with that delay, because the collagen has already become gelatin and the night has given the flavors time to move around in the liquid. The delay is a gain rather than a cost.

That is what makes the rest of it possible. One man can send out three different stews to three different tables, alone, in ten minutes.

What Willem is keeping is not a category of cuisine. It is one person’s version of it, his grandmother’s.

cocottebrugge.be

It was a warm summer evening in Bangalore, India. I was there on a work visit and decided to go for dinner with my gang of youthful friends. I ended up ordering the wine they had which was from a local vineyard not too far from Bangalore in Nandi Hills. The gentleman who poured a sample was very excited that I had ordered a local wine. But when I had taken my first sip, I felt it so harsh that I had to abandon continuing with it. For all that, I was actually a non-taster by test. This means that I am one of those who have a genetic predisposition not to taste bitterness much. This made me wonder if this was because the wine was produced in the tropics. What makes wine so special that it matters where it is grown?

Sunlight travels 150 million kilometers before it strikes the surface of earth. It is the source of energy for almost all living forms on earth. Plants are the ones harvesting that energy first through the process called photosynthesis. So do the vines that bear grapes. The vineyards and the grapes occupy a narrow band on the earth in both hemispheres. It is the zone between 30 and 50 degrees latitude north and south. To understand why, we need a flashlight and a school globe of earth.

You must have noticed that the desk earth globe shows the earth on a tilted axis. This very axis is the reason for our seasons. And the reason we have the variations that we experience at any given latitude during the seasons is the angle at which the sunlight hits the surface of the earth. Depending on your latitude, the sunlight hits you at different angles. Let us do an experiment with our desk earth globe. First we will turn the lights off in the room. We now cast the light from our flashlight onto the globe. You will notice that the bright circle of light emanating from the flashlight stretches into a dim oval. The brightness now varies based on the angle it hits the globe.

Let us consider where we are right now on earth. Even though we are on flat ground, from a perspective of the angle at which sunlight strikes us, we are on an incline. Take me, in Phoenix for example. Phoenix sits at about 33 degrees north latitude, so the flat ground I am standing on is actually tilted 33 degrees. For the sunlight that is hitting me, I am on a 33 degree hill.

Flat ground is actually an incline.

Mendoza region in Argentina is around the same latitude but on the opposite hemisphere. Mendoza produces wine but Phoenix does not. It is considered a cold desert with a similar rainfall pattern to Phoenix. But Mendoza sits at 3000 feet above sea level which adds the additional lever in lowering temperature. There is also a marginal effect from more light getting through the thinner atmosphere. In addition, the Andes provide cool winds and reliable water. All of these, with elevation being the major lever, help with the cultivation of grapes in this region.

The incline is needed but not sufficient.

Now, let us move to the upper middle of the wine belt in the old world wine growing region, Hermitage, Rhone Valley in France. Hermitage is a prestigious wine growing area located at 45 degrees north rising above the river, Rhone. Hermitage is in all about 130 hectares on one south facing granite hill with flat land all around it. The slope allows not only maximal light exposure but also protection against the mighty wind from the northern Rhone, the Mistral. The wind itself has a lever of drying out the canopy that arrests any decay from moisture. Here the slope is the biggest lever with the wind providing secondary advantage. Surrounding Hermitage is Crozes-Hermitage, eight times its size and flatter, but selling wine at a fraction of the price.

We now drift parallel to the west to the new world region, Napa valley, located at 38.3 north in California. Here the valley floor is heavily fogged three mornings a week until around half past nine during the growing season. Here you need an elevation of 1400 feet to be out of the fog layer. So all of the wine growing areas above that layer get warmer mornings and cooler afternoons along with less variation in temperature across day and night. The grapes ripen slower and later than on the floor. Elevation needs to be high enough to escape the fog.

Sauternes sits at the southwest edge of the Bordeaux growing region at 44.5 degrees north. The Ciron is a cold, spring fed river running into this warm zone. During autumn this creates morning fog which burns away into dry air by afternoons. The damp air establishes the fungus, Botrytis cinerea but the dry afternoon air prevents the rot. The fungus pierces the grape so fine that just the water evaporates concentrating the sugar up to 400g/l. The fungus also makes some compounds itself and leaves precursors which the yeasts convert during fermentation infusing flavors to this wine. My wife loves the Sauternes as she is a supertaster, the people who are genetically predisposed to be intolerant to bitterness. Here the fog makes my wife happy but it is a delicate balance. The same fungus with the wrong weather can rot the grape.

We have arrived here after we talked about incline, elevation, fog and wind. All of these can work outside of the belt as well. There are two major levers where the belt does have an upper hand. But when you combine the levers, the belt becomes the clear winner.

The first lever is the summer energy budget that gets allocated based on your latitude in summer. So let us start with allocation on one day in midsummer in the northern hemisphere and calculate how much energy is delivered at different latitudes. The equator gets the most concentrated light over the year and Tropic of Cancer has the sun directly overhead. Hence they should be getting the most energy delivered during midsummer. But that is where you have the surprise. Most of the belt beats both the equator and the Tropic of Cancer. The equator gets 20.8 megajoules per meter squared that day. Hermitage gets 26.1, a quarter more.

How is that possible? It is possible because the energy allocation is not just based on the angle at which the sun hits the latitude but for how long. During the summer the earth is tilted towards the sun. This is where we go back to the desk globe and the flashlight. The room should be made dark first. Let us start with shining the light straight with the northern hemisphere tilted towards the light and slowly spin the globe. You will see the upper latitudes stay in the light for longer and at the Arctic Circle it is all the time. The equator meanwhile receives the sunlight the same 12 hours regardless. And the Tropic of Cancer has the sun right over it in midsummer which gets it only 13 hours and 27 minutes of daylight compared to Hermitage which has 15 hours and 37 minutes.

Now let us move to the entire summer energy budget. The total amount of energy that a certain latitude gets through the entire summer is measured in gigajoules per meter squared. Hermitage for example takes 11 percent more than the equator across the whole season. Napa gets slightly more. The peak occurs at 30 degrees north at 2.32 gigajoules per meter squared. Then it sort of holds steady from 30 to 45 degrees north around 2.28 after it starts its decline to 2.15 at 50 degrees north. Now, let us talk about how this is relevant to the grapes. The flowering starts around midsummer during which the cold and rain are the enemies. The warmth from all the energy harvested helps. The growth of the vine, its canopy, leaf surface and its entire machinery occurs mostly in summer and hence the energy advantage is key.

The next lever is the most important and it has to do with the timing of ripening of the fruit. End of July is the veraison. That is when the grapes lose their chlorophyll and gain the colors. The red grapes get their anthocyanin and gain purple and red colors. The white grapes lose their chlorophyll turning golden or translucent. The skin softens and sugars start rushing into them and their acid starts dropping as we get into fall. That is when the remarkable thing happens.

The energy from the sun drops dramatically as the earth’s tilt now is away from the sun and the days shorten due to it. This is the first layer of autumn collapse of energy. At Hermitage, the daylight goes from 14 hours 42 minutes to 10 hours 52 minutes from the first of August to the middle of October. The energy falls from 23.5 megajoules per meter squared per day on the first of August to 17.4 by the middle of October from the shorter day alone. That is a 26 percent collapse at about 1.2 megajoules per fortnight correlating with forty-five minutes of daylight every fortnight without a pause.

The second layer of autumn collapse comes from the sun being lower on the horizon with the tilt. To understand this we need to go back to the flashlight. When you focus the flashlight on the wall at an angle, you will see the beam of light spread over a larger area on the wall into a pale oval. Just like that with the tilt of the globe away from the light, the same beam of sunlight has to spread over a wider area reducing the intensity of light hitting each unit area of the ground. This causes further energy loss. At Hermitage, the noon sun, which is at 63 degrees on the first of August, goes down to 35 degrees in the middle of October. The same beam now covers 1.7 times the ground in October. From the energy that is left after the first layer of loss, the second layer takes it from 17.4 megajoules per meter squared to 11.3. That is an additional 6.1 megajoules which is about 35 percent at a rate of about 1.2 megajoules per fortnight.

The third and final layer of collapse happens due to the air. We have just discussed how the earth tilts away from the sun in autumn. Because of this the light now has to travel more through the atmosphere before it touches the surface. The air reduces the intensity of light essentially by scattering of light by air itself and the particles floating in it. Additionally there is also loss from absorption by water vapor, carbon dioxide and ozone. The loss is all across the light spectrum. So when the light eventually reaches the surface there is a decrease in the energy that was left after the first two layers, 11.3 megajoules to 10.4 megajoules per meter squared at Hermitage between the first of August and the middle of October from the air alone. This is a loss of an additional 8 percent. The energy left after each stage was 0.74 then 0.65 followed by 0.92 of what came before. Multiplying them 0.74 x 0.65 x 0.92, we get a total of only 44 percent that is left, which amounts to a 56 percent loss from all the three layers of autumn collapse.

All these layers of autumn collapse do a huge favor to the grapes and more so for the wine makers. It slows down the sugar from a race to a creep. The cold nights stop the acid from burning off. Skin color matures, tannins soften and aromas start developing. It is almost as if the autumn collapse puts the grape maturity in slow motion. Sugars build slowly so they accumulate peaking at harvest. The acid burn off comes to a stop and will come in handy for preserving wine later. The character of the grape builds in the color and flavors.

This is the effect of the geometry of Earth’s tilt, sunlight and air.

There is one other factor that plays an important role in the viability of a wine growing region. It is the temperature window during the wine growing season. This is not fixed by latitude alone. For example Bordeaux and Mongolia are sitting at the same latitude with the same light and angles. What separates them is heat and heat comes from ocean, altitude and how far inland you are. Jones in 2006 determined that quality wines come from places whose growing season average sits between 13 and 21 degrees Celsius. On measured temperatures from 2016 to 2025, the 13 to 21 window runs from 32.5 to 46.8 degrees north and 30.5 to 42.4 degrees south. The temperature window then falls within the belt where light is already doing the right work with the two major levers. Every serious region sits inside this temperature zone and every one that sits outside and some that are sitting inside have other levers helping them.

Now it is time to answer our original question. Let us travel back to Nandi Hills which is at 13.2 degrees north and 45 kilometers from Bangalore. We just talked about the two big levers in wine that no one chose but were endowed: the summer budget and the autumn collapse. Nandi Hills gets exactly the same summer energy as Hermitage, 2.22 gigajoules per meter squared. Before we get to the collapse we need to talk about the one major adjustment by hand that Nandi Hills had to do. It had to change the growing season entirely due to the monsoon. The vine does not stop as the sun does not tell it to. So the vine is pruned to force a new start. The season now changed by hand at Nandi Hills runs from December to the middle of March. So the entire summer budget we talked about earlier becomes irrelevant. The difference is entirely in the collapse. Hermitage loses fifty six percent across its ripening. Nandi Hills gains twenty eight percent across its own, rising from 17.9 megajoules a day in December to 22.9 by the middle of March. In Nandi Hills, the days do not shorten, the sun does not drop toward the horizon and the air path of the sunlight does not lengthen. The grape ripens under the sun which never withdraws. They are forced to pick when the sugar is ready but the fruit is not. The acid burns off too. The pigments and the flavors do not get enough time to mature. It is an uphill battle for the wine makers here. Some of the defects are structurally built in due to the latitude. So the margin for error is negligible. No one chose these levers and the belt was not chosen either. The places that are in the belt got the summer budget and the autumn brake.

Everywhere else you have to build all that by hand.

In 1855, Napoleon wanted France’s produce shown in the Exposition Universelle. He specifically asked Bordeaux Chamber of Commerce to arrange for a wine exhibit there. The method that was used was ranking by average trade price over the previous forty years. Bordeaux sits inside the summer energy and autumn collapse belt. The brokers got it right by price and what sold. But Burgundy was refused even though it sits inside the same band.

The scaffold drew the boundary based on administrative power not based on what was grown on the ground.

This very wine belt is not a fixed entity. It has started to move. The two big levers we talked about do not move. The temperature band which decides who can grow is moving and it is because of climate change. Harvest in most of the wine growing regions is happening two to three weeks earlier compared to forty years ago. France’s harvest has moved by 24 days. A review done by van Leeuwen in 2024 finds that if global warming increases by 2 degrees Celsius, 90 percent of winegrowers in coastal regions and plains of Spain, Italy, Greece and southern California will no longer be viable at the end of the century. The winners appear to be northern France, Washington, Oregon, British Columbia, Tasmania along with new entries in Belgium, the Netherlands and Denmark. That temperature line is moving toward latitudes that have a steeper autumn collapse. The heat window is going to be a dominant player in the future.

Climate change has now created a new frontier for wine growing. This includes Hokkaido at 43 degrees, Tasmania at 42, Patagonia at 45, northern France at 50, Poland at 52, Denmark at 55. The same summer light as Bordeaux and a steeper autumn collapse. They were always within the wine belt. What kept them out was the temperature and the scaffold. Climate change is fixing the temperature.

The scaffold, where the control sits, has not moved.

The appellation needs a meritocratic upgrade and that too by physics. Science provides us a factual basis to open the gate away from the whimsical gatekeepers. 1855 ranked just what one port could see but physics admits both hemispheres. A classification drawn from the band where summer holds and the collapse runs the right way starts with what grows there not who is in the room. It has limitations as well. It can only admit places but cannot guarantee them. Nova Scotia has the light and the brake running the right way but the spring frost is the costly risk.

Physics is an honest gatekeeper. It says where the ground is right but cannot say if it is worth a name. The naming rights should belong to the folks who consume the wine. That would be the right stack. Next in the series we will explore what was actually chosen, the vine, the rootstock and the fruit.

Sources

1. Earth to Sun mean distance, 149.6 million km. Standard reference.

2. Latitudes: Phoenix 33.45°N, Mendoza 32.89°S, Hermitage 45.09°N, Napa 38.30°N, Sauternes 44.5°N, Bengaluru 12.97°N. Standard gazetteer values.

3. Day length at 45°N (15h37m and 8h46m), Tropic of Cancer (13h27m), equator (12h07m): timeanddate.com day length tables; Wikipedia, “Daytime.”

4. All insolation figures (megajoules and gigajoules per square metre, the collapse percentages, the three layers, the belt curves) are computed from standard solar geometry per Duffie, J.A. and Beckman, W.A., Solar Engineering of Thermal Processes, with clear sky atmospheric transmission per Meinel, A.B. and Meinel, M.P. (1976), Applied Solar Energy.

5. Mendoza: cold desert climate, under 9 inches of annual rain, around 70 percent of Argentine wine, vineyards near 3,000 feet. Wikipedia, “Mendoza Province” and “Mendoza wine.”

6. Hermitage: single south facing granite hill, 344 m, about 130 hectares; Crozes-Hermitage about 1,032 hectares; the Mistral from the north. Wikipedia, “Hermitage AOC” and “Crozes-Hermitage AOC”; Wine-Searcher regional pages.

7. Napa fog: inversion layer near 1,400 feet; Howell Mountain AVA (1983) defined above the fog line; valley floor fogged about three mornings a week until about 9:30. Napa Valley Vintners; Howell Mountain Vintners and Growers; TTB AVA documentation.

8. Sauternes: the Ciron meeting the Garonne; Botrytis cinerea; must up to about 400 g/L; glycerol and sotolon; preferential consumption of tartaric acid. Wikipedia, “Sauternes (wine)” and “Noble rot”; Wine Australia fact sheet on botrytis.

9. Grover Zampa, Nandi Hills: Doddaballapura, about 45 km from Bengaluru, 920 m, harvest January to March. Grover Zampa website.

10. Winkler growing degree day index, April 1 to October 31: Amerine, M.A. and Winkler, A.J. (1944), Hilgardia 15: 493 to 675.

11. Veraison at the end of July and flowering around midsummer near 45°N: Vintel viticulture guide and standard phenology references.

12. Malic acid is respired in the berry at a rate that rises with temperature; tartaric acid is stable. Sweetman, C., Sadras, V.O., Hancock, R.D., Soole, K.L. and Ford, C.M. (2014), “Metabolic effects of elevated temperature on organic acid degradation in ripening Vitis vinifera fruit,” Journal of Experimental Botany 65: 5975 to 5988. Plantevin, M. et al. (2024), “Characterization of varietal effects on the acidity and pH of grape berries for selection of varieties better adapted to climate change,” Frontiers in Plant Science 15: 1439114, DOI 10.3389/fpls.2024.1439114.

13. Jones, G.V. (2010), “Climate, Grapes, and Wine: Structure and Suitability in a Variable and Changing Climate,” VIII International Terroir Congress. The 12 to 22°C bound.

14. Jones, G.V. (2006), “Climate and Terroir: Impacts of Climate Variability and Change on Wine,” in Macqueen, R.W. and Meinert, L.D. (eds.), Fine Wine and Terroir: The Geoscience Perspective, Geoscience Canada Reprint Series 9, Geological Association of Canada, St. John’s, pp. 203 to 217. The 13 to 21°C quality band.

15. Zonal growing season temperatures 2016 to 2025: NOAA Physical Sciences Laboratory, NCEP-DOE AMIP-II Reanalysis (R-2), monthly mean 2 m air temperature. Zonal means calculated across longitude; growing season April to October in the north, October to April in the south. Kanamitsu, M., Ebisuzaki, W., Woollen, J., Yang, S.-K., Hnilo, J.J., Fiorino, M. and Potter, G.L. (2002), “NCEP-DOE AMIP-II Reanalysis (R-2),” Bulletin of the American Meteorological Society 83: 1631 to 1643, DOI 10.1175/BAMS-83-11-1631.

16. The 1855 classification: Bordeaux Wine Council, “The 1855 Classification”; Wikipedia, “Bordeaux Wine Official Classification of 1855.” Commissioned through the Bordeaux Chamber of Commerce for the Exposition Universelle; ranked by trading price; Burgundy refused; the right bank excluded.

17. van Leeuwen, C., Sgubin, G., Bois, B., Ollat, N., Swingedouw, D., Zito, S. and Gambetta, G.A. (2024), “Climate change impacts and adaptations of wine production,” Nature Reviews Earth and Environment 5: 258 to 275.

18. INRAE press release, March 2024, on the above.

19. France’s harvest about 24 days earlier; the 2026 heat; plantings in Normandy: France 24 reporting, September 2026.

20. Morales-Castilla, I., García de Cortázar-Atauri, I., Cook, B.I., Lacombe, T., Parker, A., van Leeuwen, C., Nicholas, K.A. and Wolkovich, E.M. (2020), “Diversity buffers winegrowing regions from climate change losses,” PNAS 117(6): 2864 to 2869, DOI 10.1073/pnas.1906731117.

21. Frontier regions: van Leeuwen et al. (2024) for the named winners and new entrants.

It was a warm summer morning. Me and my wife resigned ourselves to having an upper body workout in our backyard. We made a pact to split the effort by taking turns. It was going to be worth it. You might be wondering what am I talking about. Believe it or not, I grow sugarcane in my back yard. Plenty of sun and a modest water bill makes it happen here in Arizona. My wife’s nostalgia arising from quenching her thirst in the humid bustling city of her youth along with my own younger days on my grandfather’s sugarcane farm make this a joint voluntary effort. I am talking about manually crushing sugarcane for that juice delight.

We squeeze a sliced piece of ginger and lime every other crush of our carefully cut sugarcane stems. After an hour or so of intense workout, the job is done. We have two big jars of this fresh juice with an olive hue. The sugarcane we grow is the Mexican variety. Sugarcane stores its sugars in the form of sucrose which by itself is sweet on your tongue. The Mexican variety has lower levels than the ones we both grew up with in India.

The clock starts ticking once the sugarcane is cut. The unseen hands of the ancient cooks start to act. Sugarcane inherently has an enzyme called invertase which starts breaking down sucrose into fructose and glucose.[1] Besides, we also have microorganisms coming to feast with the same ammunition. Those jars of juice are going to be their playground.

The clock starts ticking once the sugarcane is cut. The unseen hands of the ancient cooks start to act.

My wife and I then eagerly move those jars straight into our refrigerator. The olive colored juice mildly corrupted with juices of lime and ginger is in the custody of a refrigerator. While the microorganisms are converting the sucrose, the acidity created by lime encourages the invertase to work even more. Here we have yet another clock. The race for the conversion to the simpler sugars. Too fast and the juice will start to sour as the simpler sugars will be eaten. This is where the refrigerator lends its hand. The colder temperature slows down the microorganisms enough to be able to store the juice for a couple of weeks. We literally slow the clock down. In the meanwhile the juice gets sweeter every day because of the simple sugars accumulating. That is the summer delight we both are after.

We literally slow the clock down. In the meanwhile the juice gets sweeter every day.

Back on my grandfather’s farm, the sugarcanes will be crushed right after harvest. The juice then will be poured into this giant cauldron heated by wood or biogas. That is when the magical journey will start for a young kid I was then. The juice will undergo a series of transformations that I will be waiting to witness and experience with my greedy palate.

The juice starts turning into darker shades of brown through this journey. First it will turn into this thick syrup which will be the right opportunity for me to dip the tender coconut pieces in to come out as this toasted coconut delight. Then the syrup will transform into this thick jelly and I will be waiting to dip my stick in to make an amazing candy on a stick. Eventually it will become this brown powdery jaggery which will be molded into these mini pyramids. Part of that I know will end up in my grandma’s kitchen to become memorable desserts. That is the impact sugarcanes had on my childhood.

In the cauldron, a different ancient cook takes over. Heat takes out the microorganisms when the temperature climbs beyond 60 degrees Celsius. The heat by itself does the conversion now faster than any enzyme could do. The conversion is done by the new cook through a purely chemical process. As the water in the juice boils off, the boiling point gets reset higher with it. Every stage I was experiencing through my palate was a concentration reading in disguise. Thin syrup toasting the coconut happens around 105 degrees Celsius. The candy on stick happens at around 110 degrees Celsius. The jaggery stage occurs around 118 to 120 degrees Celsius while being stirred which seeds thousands of tiny crystals making it a fine powder to be compacted as pyramids.[2] Glucose and fructose are reducing sugars and react with amino acids in the cauldron triggering a Maillard reaction once the temperature crosses 110 degrees Celsius. Caramelization adds on from the heated surface of the cauldron which enhances the browning. The sugar mill stops well before this conversion and crystallizes sucrose packing it off.

Every stage I was experiencing through my palate was a concentration reading in disguise.

Fermentation is the most ancient cook here. A fermented fruit on a forest floor digested by microorganisms produces ethanol which is then converted into energy by the animals that consume them aided by the enzyme alcohol dehydrogenase (ADH). This is a process of digestion outsourced which improves the energy economy of the animal feeding on it by conserving the energy spent on digestion. The twist to this evolutionary tale is that around 10 million years ago, in one of our early primate ancestors before we broke from chimps and gorillas, a single amino acid change in the ADH4 enzyme made ethanol metabolism roughly 40 times more efficient.[3]

This is a process of digestion outsourced.

This mutation became a game changer for the energy economy of these animals. Fermentation became the first pre-digestion technology.[4] There is a clock here too, the fermented fruits have to be picked within a certain time exceeding which it will sour. This technology became an effective and efficient strategy in energy harvesting. So much so, early humans have one of the smallest guts for an ape of our size.[5] Anatomical adaptation only occurs after prolonged adoption of outsourced digestion.

Fire as a consistent pre-digestion tool emerged much later around 400,000 years ago in our human ancestors.[6] It could have been much earlier when we or our ancestors used wild fire but that would have been too sporadic to be adopted as a technique.

The first meal that would have been cooked with fire is unknown. The earliest evidence points towards fish.[7] Meat probably followed but the best energy return belongs to starch.

In the case of proteins in meat, heat unfolds them and opens them to proteases in our guts. Collagen, which was untouchable before, melts to gelatin. Now it is cheaper to extract calories from meat. Meat is also easier to chew. Energy gain compounded.

Starch on the other hand is stored in granular form which is nearly impenetrable by amylase, an enzyme in our guts. Heat makes the starch swell up and the granules burst making it now digestible by amylase. This turned tubers and roots from fillers into energy bursts. Our human ancestors had used mechanical pounding to unlock some of the energy release before.[8] But fire improved the energy gain by about a third.[9] Besides, there is again some compounding from the savings of avoiding the mechanical pounding.

Heat also made these food sources safer by neutralizing toxins and made meat last longer. Some plants became food only after fire. Besides unlocking these food sources, heat also gave us two other enhancements. It returned us the hours on chewing and pounding that we could use the big brains for.[10] The heat induced Maillard reaction and caramelization produced molecules that no biology can make. The flavor profile these created stays with us to this very day.

There has been much debate whether fermentation or fire contributed more to our brain expansion. But I do not believe it is a contest at all. Fermentation emerged early as an advancement in energy economy and fire came later as an additional layer on the stack to give a further energy boost.

Fermentation vs Fire: But I do not believe it is a contest at all.

Our cultures have recognized this non contest from time immemorial. It is visible once you expose these ancient cooks in our everyday dishes.

The first among those dishes is the quintessential bread. The making of the dish starts with the prep cook - the mechanical force in milling the grain to flour and kneading it to dough.

After the dough is kneaded it is passed onto the in vitro cook station where the yeast crew works on the dough from within.

Let us now dive into how the crew works inside the dough. The work is actually fun because it involves eating the sugar in the dough. All the eating makes them release carbon dioxide as an exhaust. Thanks to the prep cook, the kneading has built the gluten into a stretchy net trapping the exhaled carbon dioxide in thousands of tiny balloons. This causes the rise in the bread eventually. Besides this physical impact, the crew’s eating, helped by the dough’s own enzymes, breaks down the long starches into simple sugars easing the digestive burden on the consumer. The predigestion is thus enacted. The crew also throws off alcohols, esters and organic acids adding an entire aroma cabinet which we perceive as the taste of bread. Add some bacteria and you get the sour in the sourdough bread.

Next, it is the turn of the line cook, the thermal cook. Even though fire was our ancient thermal cook, we have since then shifted the source to many including the oven. The dough is shifted to the oven. The crew is still alive in the dough up to 60 degrees Celsius. Actually as the temperature climbs, the crew goes into hyperdrive until they all perish. This builds up more sugar that contributes to the browning and the gas that gives that extra lift. Therefore we need the temperature climb where preheating the oven becomes very handy. This accelerates the climb as the dough is a poor conductor of heat and the climb inside needs the help of the oven blast.

So the thermal cook converts the foamy dough into a sponge of inflated balloons, bursts the starch granules exposing the inside for our amylase, browns with the crew’s leftover work and spills the aroma cabinet through the steam escaping the pores.

During my recent trip to the Amazon, me and my family visited an Indigenous settlement. There we saw them grate cassava, or manioc as they call it. Grating makes the plant’s own enzyme release a deadly poison, hydrogen cyanide. Then they wash it repeatedly followed by fermentation and drying getting rid of the poison.[11] The edible powder toasted over fire is called farinha, the manioc meal. We had that meal, so we know it is poison free. The washed liquid, the tribe leader told us, they paint on their arrowheads and use to hunt wild pigs.

So we see the prep cook with grating and washing, in vitro cook with fermenting and the line cook with toasting. They are all working in sequence. Any switch in sequence, like putting the line cook first, can be deadly. The waste product is put to use here to gain another energy source efficiently. You can see the energy economy compounding.

Any switch in sequence, like putting the line cook first, can be deadly.

We are circling back to my home for the next dish. It is idli, about whose history I wrote an essay not too long ago. Here we have both rice and black gram mechanically ground with water into a liquid batter. It is then left overnight to ferment.[12] Eventually the batter rises after several hours of our in vitro cook working. Then the batter is poured onto circular moulds in a steam cooker with water at the bottom. So we have the prep cook creating the batter and pouring into the moulds. The in vitro cook does the same work with the crew like in the case of the bread. The thermal cook acts through steam which arises by heating the water in the steam cooker with fire. The steam when it hits the cool batter condenses on it as water, releasing heat at a constant 100 degrees Celsius which becomes its ceiling. The wet batter conducts the heat inwards faster and gives all the water for gelatinization setting idli up for a softer center instead of crusted as in bread. There is a climb inside of the batter which does give an uplift to the rise. The thermal cook as with the bread converts the foam into sponge. The ceiling of 100 degrees Celsius does not allow for Maillard browning which needs 140 degrees Celsius and so the batter turns into these beautiful soft white cakes. Maillard kicks in when you pour the same batter on a hot pan to make dosas (crepes). The thermal cook here works through a nuanced process to deliver a temperature held constant which helped shape the dish.

The last dish for this sequence is Miso which interestingly will inverse the sequence of the ancient cooks. The line cook gets recruited first to boil the soybeans to soften the impenetrable wall and disarm the inhibitors of our enzymes. Then the in vitro cook is introduced in the form of koji which is cooked rice inoculated with a domesticated aspergillus, Aspergillus oryzae.[13] This introduction is engineered by the prep cook by mashing the boiled beans with koji and salt at 5 to 13 percent. That salt behaves like a bouncer selectively recruiting salt loving yeasts and lactic bacteria. Meanwhile the enzymes of the mold continue to shred the proteins in the soybean into amino acids even long after the demise of the fungus. This mishmash of goodness is fermented for months as a reservoir of umami for a variety of dishes.

We have now looked at all these dishes where the energy stacking and compounding is obvious. The jug which banks the sweetness, the cauldron that packs the calories into portable pyramids, bread and idli converting grains into filling nourishing sponges, farinha turning poison to food along with the poison to pigs and finally miso converting beans into a mobile amino acid reservoir. From the onset of ADH4 mutation, the ancient cooks are doing the same errand of energy surplus. More energy harvest with less energy spent. You also cannot help noticing all these dishes are rather humble for people with limited resources. They are either street, farm or peasant food. They are for poor people with full kitchens and armored foods. They did what evolution offered, recruiting ancient cooks, the natural pre-digestion technologies to run an energy surplus as their resources are scarce.

If you want to measure true sophistication in a kitchen by the number of ancient cooks deployed and not by the price charged, these kitchens come out first. The priciest kitchens often use the most unarmored food and the fewest of these ancient cooks calling themselves sophisticated.

With thanks to @Oscar Waters, whose seeds grew into this series, and to @Andrew Strauss, whose observation about the temperature climb became this essay’s schedule thread.

Notes

[1] Solomon S, Post-harvest deterioration of sugarcane, Sugar Tech 11(2), 2009.
[2] Jagannadha Rao PVK et al., Jaggery, a traditional Indian sweetener, 2007.
[3] Carrigan MA et al., Hominids adapted to metabolize ethanol long before human-directed fermentation, PNAS 112(2), 2015.
[4] Bryant KL, Hansen C and Hecht EE, Fermentation technology as a driver of human brain expansion, Communications Biology 6:1190, 2023.
[5] Aiello LC and Wheeler P, The expensive-tissue hypothesis, Current Anthropology 36(2), 1995.
[6] Roebroeks W and Villa P, On the earliest evidence for habitual use of fire in Europe, PNAS 108(13), 2011.
[7] Zohar I et al., Evidence for the cooking of fish 780,000 years ago at Gesher Benot Ya’aqov, Nature Ecology and Evolution 6, 2022.
[8] Zink KD and Lieberman DE, Impact of meat and Lower Palaeolithic food processing techniques on chewing, Nature 531, 2016.
[9] Carmody RN, Weintraub GS and Wrangham RW, Energetic consequences of thermal and nonthermal food processing, PNAS 108(48), 2011.
[10] Organ C et al., Phylogenetic rate shifts in feeding time during the evolution of Homo, PNAS 108(35), 2011.
[11] Montagnac JA, Davis CR and Tanumihardjo SA, Processing techniques to reduce toxicity of cassava, Comprehensive Reviews in Food Science and Food Safety 8, 2009.
[12] Steinkraus KH et al., Studies on idli, 1967.
[13] Gibbons JG et al., The evolutionary imprint of domestication on the genome of Aspergillus oryzae, Current Biology 22(15), 2012.
General kitchen-science reference throughout: McGee H, On Food and Cooking, revised edition 2004.

I was scrolling through my Substack app and there it was; one more note which said “If you have less than 100 subscribers, I want to subscribe to every one of you. Drop me a message regarding what you write about.” There were already 300 replies. I got suddenly hit with FOMO and plugged my thing immediately as a response. It felt so generous that it felt stupid not to respond. Then I started noticing something strange. Quite a few of these accounts had thousands of subscribers. Amongst them there were also folks with paucity of subscribers who were trying to connect and network emulating that approach. I started scratching my head trying to figure out why these Substackers with thousands of subscribers were throwing these hooks to connect to random people. I noticed a pattern of these engagement notes from these folks: they were either generic questions designed purely as a hook or outright offers to subscribe to anyone who connects. I realized that besides the meaningful handshake and networking, there was something else at play here.

Substack is a network of people who write and people who read. To see what a network does, it helps to look at one whose shape we already know from the inside- Airports.

The airport network starts with the major hubs like Atlanta, Chicago, Dallas-Fort Worth. They have more routes because they already had more routes. The airlines add a new route where connections already exist, creating multiple one-stop journeys from one new route. There was no plan for them to be hubs. They were larger to start with and just compounded. The slots are the scarce commodity at the hub. There are only a few runways and so many landings an hour, so the airlines that have held them for long enough continue to hold them. Next in the network come the small hubs like Salt Lake City, Minneapolis, Charlotte. They became hubs because a major airline chose them, not because a city demanded it. They can lose their hub the moment the airline decides, as it happened to Pittsburgh.

So for a small hub, the strength is borrowed.

Following the small hubs, the network has the in-betweeners. Mid-sized cities like Raleigh, Columbus, St. Louis. They get reasonable service but never originate anything. Every route is a spoke to a small or large hub. They are the majority of airports that a traveler uses. Then, last in the network, come the small nodes. The regional strips like Lubbock, Bozeman, Flagstaff, Erie have one or two routes to the nearest hub. They have service at all because of that one hub. But the small node gets the world through that hub.

Airport map: The same forty-six cities. Every line runs both ways: the hub gives the network back what it takes.

Now back to Substack. The host with thousands of subscribers is the hub; the 300 folks who are replying are the small nodes. The hub compounds with the cloak of generosity, with 300 new connections. The node got one connection, with the hub routing nothing back. The small node at the airport gets the world through the hub.

The small node in the reply thread on Substack gets just the feeling of the world.

Substack map: The same cities, the same shape. Every arrow points inward. Nothing routes back.

If the small nodes did not get much, why do they keep responding? They are handed a belief and the belief does the work. The belief is that you have to put yourself out there to be found. Reply, engage, comment, restack: you are pounded with this, mostly by hubs. Participation is the price of visibility. If you reply and you get subscribers, participation worked. If you reply and nothing happens, you did not participate enough. What is never made visible is that the outcome was set by where you sit in the network. The one writer out of 300 who got discovered sells the belief for free.

This is the visibility rig.

The Michelin system works in similar ways. Here it starts with the admission rig, whereas there is no gate on Substack. The admission rig is granted in return for payment. Virginia refused to pay $360,000. Atlanta paid $300,000 a year to be admitted, with annual renewal set at $225,000 a year. Boston paid $1 million over three years for 26 restaurants and one star, and the hotel guests are paying for that through their room surcharge. Minneapolis paid $250,000 a year, and the coverage there stops at the city line because the tourism district that paid for it stops there.

Next, Michelin hands us the second admission rig. The inspectors do not eat everywhere.

They eat where they are pointed, and even among those, they choose.

Nobody is told why, or by what method. Emails obtained by Axios show the tourism boards nominating the restaurants they want visited. A minority-cuisine restaurant that no tourism board put on its list will never be visited.

Michelin map: The same map a third time. Fourteen cities lit because someone paid. The rest were never visited.

Every network has a gate. Airports have gates as well, for the traveler at the node level, but they open for a cost and keep it transparent: the airport charges, but it charges everyone the same and sells something real. Substack has a gate at visibility, not at admission, and no transparency. Michelin gates you at admission and at visibility.

Two gates become a moat.

Besides the major hubs, which are mostly too big to be restructured, the rest of the airports reflect the cities they sit in and how those cities are endowed. It is demographics, terrain, proximity to a bigger hub, economy, diversity of travel habits. But talent, like writing or cooking, has no geography.

So, a network concentrating writers or cooks is not reflecting an endowment, it is manufacturing one.

These are real defects: networks rigged against their own users. On Substack, visibility is rigged. Position in the network sets the odds, and a belief keeps you playing. At Michelin, both admission and visibility are rigged. A city pays to exist, and a tourism board and the inspector decide who gets judged at all. Both call the result merit. The talent was always distributed evenly. The counting never was.

Michelin is one opinion, never moderated, never aggregated. That’s not judgment. That’s a pronouncement.

Sources

Virginia’s refusal and the $360,000 figure: Axios Richmond, April 2025. Atlanta’s $300,000-a-year contract, the $225,000 renewal, and the emails showing tourism boards submitting restaurant lists: Axios Atlanta, “Atlanta wasn’t sold on a Southern Michelin Guide, emails show,” May 23, 2025. Boston’s three-year deal of just over $1 million, yielding 26 restaurants and one star: The Boston Globe, reported via Nexstar, April 2026. Boston’s hotel-room surcharge as the funding source: Food Drink Life, “Your hotel bill might be funding a Michelin star,” July 16, 2026. Minneapolis’s $250,000 a year and the city-limits boundary: Axios Twin Cities, April 9, 2026; Minnesota Monthly, April 8, 2026; Star Tribune, April 8 and June 23, 2026. Pittsburgh’s loss of hub status: US Airways dehubbing, 2004.

It happened again. I woke up incompletely with this pounding headache. But somehow my sleep wrestled it down to numbness until it could not anymore. I woke up finally with this vague listlessness. My mind was unfocused and wanting not to be taxed with any thought. Fighting through that fog, I started to realize what was wrong. I immediately made hot coffee with a generous amount of sugar. Through all of this, I was feeling this vague sense of nausea and hunger mixed together. Despite that, I started guzzling lots of water like I was in a desert on a hot day. Of one thing I was certain, this too shall pass.

What I served you here is my version of a hangover. It could have been the name of a particularly hilarious movie series, but this is not a funny experience for me. I remember vaguely going through this during my much younger days of reckless binges and endless parties. But lately, as I am getting older, I am starting to notice this coming back. I am more thoughtful and disciplined, as a drinker at least at this age. So, I have decided to take the journey of discovering what happens to alcohol after it passes my palate.

Before we proceed further in this journey, let us meet and greet the alcohol we consume, ethanol. It is important because every other alcohol is not as friendly. Ethanol is a small molecule with two carbons and has some cool properties. It is water-soluble and fat-soluble. Sort of a dual citizenship. Our membranes are fat, blood is water, and it is at home with both. Ethanol is not only consumed with food but is made from food as well by a process called fermentation where food is eaten by yeast and ethanol is exhaled. By food I mean grape, grain, rice, honey, most of which you will find in your kitchen and pantry.

After drenching my senses and slipping through the food pipe, ethanol lands at its first stop, the stomach. The fascinating thing about ethanol is that it needs no digestion at all. Everything in the stomach needs to be broken before the body can use it but ethanol crosses by passive diffusion. It is like a ghost walking through doors. No transporter, no carrier and nothing in the stomach lumen competing with it to enter. It comes ready to walk through the moment it is consumed.

The stomach mostly functions as a custodian rather than an absorber. Some alcohol does get absorbed through the stomach walking through as it does everywhere. But the stomach is a small smooth sac which makes its surface area finite. The blood supply is also not very rich and therefore even the alcohol that walks through has a traffic congestion issue. This makes alcohol unable to cross the wall and look for the exit through the valve, the pylorus, at the lower end of the stomach which acts as the gatekeeper.

The next station is the duodenum, the part of the small intestine which comes right after the stomach. The key to this door is on the side of the duodenum. The special sensors it has detect the nutrients it has already received and tell the stomach to wait until it is ready for some more. This feedback causes the flow to slow down to a trickle enabling efficient absorption of nutrients. So the wisdom to line your stomach with food while drinking is the right advice. But the power broker here is the duodenum who holds the lever. This slow trickle in turn lets only a trickle of alcohol into the duodenum where, aided by large surface area and rich blood supply, almost all of it is absorbed as soon as it arrives.

Once ethanol walks through the door, regardless of whether it is stomach, duodenum or further down, it ends in the portal vein, which is the direct route to the liver. So, contrary to the popular accounts, there is no alternative route that skips the liver. There is no back door or back alley. Every drop has to check in with the liver.

So what happens to ethanol in the liver? The liver has an enzyme called alcohol dehydrogenase, popularly known as ADH. This enzyme metabolizes ethanol but has a saturation point. This saturation point is set at 15 to 20 mg per 100 mL, which is roughly a quarter of the legal driving limit. [1] Beyond that saturation point the liver machinery cannot handle any more alcohol. The enzyme cannot be made to process more alcohol by practice, tolerance or experience.

That is a ceiling that does not move.

If the alcohol arrives slower than the ceiling it is processed on first pass. But if ethanol comes as a bolus way above the ceiling, the excess just gets past the desk into the general circulation. That is why the gate upstream matters.

The valve is the only variable in the system. Everything downstream is fixed.

There is a second reason food helps handle ethanol better. This one is even more of a surprise. We saw earlier that food helps hold the gate longer in the stomach. Food can help the liver eliminate ethanol 25 to 45 percent faster than drinking on an empty stomach. [2] How can that happen when the ceiling is fixed? That is the twist in this tale.

The key to this twist is that the enzyme, ADH, needs a helper by the name of NAD (nicotinamide adenine dinucleotide). For every ethanol molecule ADH processes it needs one molecule of NAD. This molecule helps absorb the hydrogen that gets stripped out of ethanol. NAD becomes NADH. The more NAD the liver generates, the less time ethanol has to wait in line to be processed.

Coming back to the twist, a meal increases blood flow to the liver. The increased blood flow to the liver has a direct impact. The increased blood flow presents more alcohol to the liver. A liver with nutrition repleted, in a well fed state, regenerates and recycles NAD faster. This helps the liver process ethanol faster.

Velocity increases, not capacity.

It does not move the ceiling higher because it cannot. But this helps ADH reach the ceiling faster. So, now we know how food actually impacts the processing of ethanol. It is not for all the reasons you have been told before. It is the valve, the circulation and the nutrition. Now, you know the real story.

Now we land back with my problem. This annoying brain fog and the pounding headache. What can possibly make that happen? There have been a lot of folk tales and quackery that have tried to explain this away. But we need to know the real story. Acetaldehyde, which is one of the products of ethanol metabolism, has been the usual villain in these stories. But it is dead on arrival as acetaldehyde does not cross the blood brain barrier. [3] Dehydration has been refuted by studies and certainly cannot explain all the symptoms. [4] What actually has been indicted by science are inflammatory agents like IL-6, TNF-alpha, CRP. [5]

The morning after is an immune event, not a poisoning or water depletion.

A study was done on a group of people who were equally split between those who get hangovers and those who do not. The same amount, 11 alcoholic drinks, was consumed and they reached the same estimated alcohol level, 0.18 percent. [6] The folks who are prone to hangover reported more severe symptoms mostly cognitive, while the resistant group reported less severe symptoms and mostly being tired. The usual suspects were held the same and the groups had a different aftermath. The study did not account for how fast the alcohol levels were achieved and whether food was consumed with the alcohol. So there may have been other factors that could have affected the outcome.

Reaching for coffee, loading it with sugar, drenching my gut with water, these were all folk remedies I turned to out of desperation. In fact, an over the counter anti-inflammatory like ibuprofen would have been the more rational choice. [7] Data also suggests that as we grow older the hangover softens. [8] We do get more measured in the amount we consume and do not have as many binges. There are other factors to consider as we age like being on long term medications that can impact the hangover in a negative way. The common ones are the good old aspirin, [9] H2 blockers prescribed for acid reflux [10] and GLP-1 drugs which are fairly common these days. [11]

If the hangover was an immune event caused by inflammatory agents, that cannot be good news for our heart. Heart disease, especially involving blockage of its arteries, is strongly linked to inflammation. So, what happens to the narrative that drinking in moderation protects your heart? The observational data actually shows that moderate drinking decreases the inflammatory markers. [12] But hangovers will tend to undo that. At moderate drinking there is no conclusive evidence that ethanol either causes heart disease or protects it. [13]

So what about wine, and that chemical in it, resveratrol? The studies conclude that neither wine itself nor the chemical has any conclusive evidence to support a cardioprotective effect. The wine versus non-drinker comparisons have been contaminated by demographic selection bias such as socioeconomic status and healthcare access, along with the abstainers who tend to be a sicker population to compare with. The Chianti study, where resveratrol levels were measured in the same population of wine drinkers with very similar lifestyles, found no evidence of the chemical impacting inflammation or heart disease. [14] The wine industry has found the narrative convenient and has woven it into its marketing.

Let us get back to my problem and my investigation. I did some retrospective data mining into my hangover episodes. I tracked back my last few hangovers and the associated drinks that perpetrated them. I found an interesting fact. I have been inadvertently consuming drinks with higher ABV just because of my tendency to be creative with cocktails or preferring wines that tend to have higher ABV where the labels can be deceptive. Basically not being very mindful about it. It all made sense now.

There is another twist in my investigation. I am one of those rare late-onset migraine sufferers. There is evidence to support that people like me are not more prone to hangovers but just to having headaches as the predominant symptom. [16] That also made sense to me. There is also no strong evidence that red wine specifically triggers migraine. [15] So, avoiding red wine is not warranted. Certainly, my life is an example of that. I actively drink red wine and have a healthy wine cellar. Migraines have been an equal opportunity provider in that regard.

My verdict: drink less, with food, lower ABV. Not as a prescription, but for the pleasure of it.

Notes

1. Hepatic ADH saturates above 15 to 20 mg per 100 ML. Jones AW, “Alcohol, its absorption, distribution, metabolism, and excretion in the body and pharmacokinetic calculations,” WIREs Forensic Science (2019).

2. Food raises the elimination rate 25 to 45 percent. Ramchandani VA, Kwo PY, Li, Journal of Clinical Pharmacology 41:1345 (2001), using intravenous clamping so the alcohol never passes the stomach. Jones AW and Jonsson KA, Journal of Forensic Sciences 39:1084 (1994), found 36 to 50 percent, and the effect held when the meal was eaten five hours after drinking.

3. Acetaldehyde does not cross the blood brain barrier. Iseric et al., Alcoholism: Clinical and Experimental Research (2024).

4. Dehydration markers, including vasopressin, showed no correlation with hangover severity. Penning R et al., Current Drug Abuse Reviews 3(2):68 (2010).

5. Hangover severity correlates with IL-6, TNF-alpha and CRP, and at four hours it is blood ethanol rather than acetaldehyde that tracks IL-6. van de Loo AJAE et al., Journal of Clinical Medicine 9(7):2081 (2020).

6. Hangover-prone and hangover-resistant drinkers matched at roughly 11 drinks and an estimated 0.18 percent. van de Loo AJAE et al. A companion study found no cytokine differences between the two groups, so the cause of the divergence remains unknown.

7. No hangover remedy has good evidence behind it. Pittler MH, Verster JC, Ernst E, BMJ 331:1515 (2005), eight randomized trials. Roberts E et al., Addiction (2021), twenty-one placebo-controlled trials, all of very low quality. Neither paracetamol nor aspirin has ever been tested in a placebo-controlled hangover trial.

8. Hangovers soften with age. Tolstrup JS et al., Alcoholism: Clinical and Experimental Research 38(2):466 (2014), 51,645 Danes, with the effect not explained by usual amount, binge frequency, or the proportion of alcohol consumed with meals. Verster JC et al., Alcohol and Alcoholism 56(5):589 (2021), 761 Dutch drinkers, severity declining with age even after correcting for estimated blood alcohol.

9. Aspirin raises blood alcohol after ingestion. Roine R et al., JAMA 264:2406 (1990). It both speeds gastric emptying and inhibits alcohol dehydrogenase.

10. H2 blockers inhibit gastric alcohol dehydrogenase and raised absorption of small morning doses by more than 60 percent, but the effect was greatly attenuated before dinner. Palmer RH et al., American Journal of Gastroenterology 86:1749 (1991). Under real-life drinking conditions the effect disappeared. Brown AStJM and James OFW, Alimentary Pharmacology and Therapeutics 12:141 (1998).

11. GLP-1 drugs are unresolved. Participants taking them showed a slower rise in breath alcohol and felt less intoxicated (Scientific Reports, 2025), while a separate study found the drugs reduce the enzyme that metabolises alcohol and produce higher blood alcohol that takes longer to fall (npj Metabolic Health and Disease, 2025). Mouse work found no effect at all (JCI Insight, 2023). Nobody has measured the hangover.

12. Moderate drinkers show lower inflammatory markers in observational data. Bell S et al., Journal of Internal Medicine (2017), Whitehall II, 8,209 participants. Volpato S et al., Circulation 109:607 (2004), 2,574 adults aged 70 to 79.

13. The cardioprotection question turns on study design. Pooled Mendelian randomization data show no association with ischemic heart disease where cohort data show benefit (Nature Communications, 2024). Correcting for abstainer and sick-quitter bias in one cohort reversed the finding into a dose-response relationship with coronary artery calcium (European Journal of Preventive Cardiology, 2026). The American Heart Association’s 2025 scientific statement concludes that whether moderate drinking lowers risk cannot be answered at the current level of evidence.

14. Resveratrol levels were not associated with inflammatory markers, cardiovascular disease, cancer or mortality across nine years in 783 adults in the Chianti region. Semba RD et al., JAMA Internal Medicine 174(7):1077 (2014).

15. Red wine is named as a trigger by 77.8 percent of migraine patients but consistently produces an attack in only 8.8 percent. Onderwater GLJ et al., European Journal of Neurology 26(4):588 (2019). Prospective studies limit its importance considerably, and migraine patients drink less than controls. Panconesi A, Journal of Headache and Pain 9(1):19 (2008).

16. Migraine sufferers are more vulnerable to migraine-like hangover symptoms, though not to other hangover symptoms. Zlotnik Y, Plakht Y, Aven A, Engel Y, Bar Am N, Ifergane G, “Alcohol consumption and hangover patterns among migraine sufferers,” Journal of Neurosciences in Rural Practice 5(2):128 (2014), 692 Israeli university students of whom 95 met criteria for episodic migraine. A Japanese cross-sectional study of 2,577 headache sufferers found women with migraine more susceptible to hangover than women with other headaches, among non-flushers only. Yokoyama M et al., Journal of Headache and Pain 13(2):137 (2012).

I was beckoned by my son who had just entered medical school for some home made food. I know I spoil him but more importantly I love to cook for him. But this time I also had an unusual addition, my friend’s daughter who was healing from sports injury also requested the same as she could not cook it herself. I then realized what was going on here. Both were under some stress and the best therapy is some comfort food. Both these kids are the members of the Indian diaspora who have an affinity to this particular comfort dish. We are going to explore the journey of this dish back in time.

Before we start this journey, we have to sort out a confusion. Is Butter chicken another name of the same dish or is it its own? It is in fact a different dish but almost a sibling. Butter chicken is a star in its own right. It is still one of the most googled Indian dishes in the world. Before we go further into this family history we need to travel back to the common ancestor.

The common ancestor is chicken tikka. The cooking utensil for cooking this dish is classically, the tandoor, whose own history we’ll come to. The tandoor is a clay pot which is wide in the middle, narrow at both ends, the top and the bottom. The top has an opening, and the bottom is closed. The bottom holds, typically, burning charcoal. It’s the source of fire and heat. The clay wall can trap the heat inside capable of generating temperatures as high as 600-800 degrees Fahrenheit. In the tandoor, the chicken is held vertically pierced on a stick or a rod, which is kept standing on top of a fire. As the chicken cooks, the fat from the chicken melts and hits the fire. The smoke that comes out of it breathes back into the chicken, giving it this wonderful smoky flavor and aroma.

There are other chicken dishes that are cooked in the tandoor. A close sibling of chicken tikka is tandoori chicken. The interesting fact is that the marinade used for both is almost identical. The only difference between tandoori chicken and chicken tikka is that tandoori chicken is bone-in chicken cooked in large pieces, and chicken tikka is a boneless chicken, which is in the form of small cubes. The other difference between the two is that there’s only one marination for chicken tikka, but in tandoori chicken, after the first marination, before it’s put on the grill, spiced yogurt is applied on top all around the chicken. Also, the amount of acidity that’s added to the marinade is higher in tandoori chicken. The increased acidity in the marinade is to mitigate drying of the tandoori chicken as it is typically cooked longer at higher temperatures charring the edges. In the case of chicken tikka, lower acidity in the marinade suffices as they are typically cooked at a little bit lower temperature resulting in a meat that tends to be softer and not charred. Tandoori chicken is dark in color and intensity, while the chicken tikka is more in the shades of deep orange.

Tandoori chicken gets a pigment layer chicken tikka never receives, then chars on top of it.

The vessel tandoor has been in the Indian Subcontinent, dating back to the Harappan times. Tandoor-style ovens at Kalibangan and Harappa have been discovered and dated to as far back as 2600 BCE, with the wider clay oven tradition going back to the Ceramic Neolithic in the Near East, 7th to the 6th millennium BCE. The name arrived later, along the Persian road. The modern word tandūr comes from Persian tanūr, which itself came from the Akkadian tinūru, mud and fire.

Let us step back in time again for a bit. Legend has it that the Moghul emperor Babur insisted on eating boneless chicken pieces to avoid choking on bones, establishing the concept of grilling boneless meat chunks on skewers over open fires. The Urdu word tikka comes from the Turkic/Persian tekke, meaning “bits” or “pieces”. To stop the boneless meat from drying out over open fire, they marinated the pieces in yogurt, herbs, and mild spices drawing from traditional Persian meat-marinating techniques. The evidence supports the kebab lineage, which is attributed to Central Asia and Persia. Tandoori chicken was the original inhabitant of the tandoor created originally in modern-day Pakistan before partition in Peshawar by a single family that would later lay claim to it. Modern-day chicken tikka was the subsequent adaptation by the same family after the partition. So the older version of the dish got a new coating and a vessel but kept its name.

Now we have it: the vessel, the meat and the coating, each carrying a history of its own.

But the meat and the coating have a history layer that is much richer and deeper. The meat itself the way it is cooked has travelled through many a creature. Let us go back to our ancient text, Susruta Samhita, documented around 300 CE, a meat dish, Kandupachita is specifically dipped in mustard oil and powdered aromatics, then roasted over charcoal to a honey colour. In Sangam Literature, Purananuru, around second century CE, there is a recipe written for goat coated with spices and roasted over a flame. In that same period in Sangam Literature, Porunarāṟṟuppaṭai mentions sheep roasted on iron skewers. Later, in 1130 CE, comes the Manasollasa, which we probed in our Cake story in the past. The same text has a detailed recipe, Sulya Mamsa, where raw meat is coated with spices including ginger, fenugreek, coriander and salt and roasted over a flame.

विध्वा तु बहुशस्तानि बहुरन्ध्राणि कारयेत् ।
हिङ्ग्वार्द्रकरसैर्युक्तं सैन्धवेन च पेषयेत् ॥
शूलप्रोतानि कृत्वा तान्यङ्गारेषु प्रतापयेत् ॥

Vidhvā tu bahuśas tāni bahurandhrāṇi kārayet | Hiṅgvārdrakarasair yuktaṁ saindhavena ca peṣayet || Śūlaprotāni kṛtvā tāny aṅgāreṣu pratāpayet ||

Pierce them many times over, so that many holes are made. Grind them together with hing and the juice of ginger, and with rock salt. Then, having impaled them on skewers, roast them over the embers.

Mānasollāsa, third Viṃśati, Annabhoga, verses in the 1460s

Yogurt was used in meat stews and is mentioned in parallel in Indian texts dating way back. The lack of recipe does not rule out its use as marinade. Nor does its later use in Persia make them the inventors. What we now established is the long tradition of spiced meat skewered on iron rods directly on ember and over a fire. Ancient India also had tandoor like ovens and lack of written accounts does not rule out the use of them as well in roasting meat.

The most important ingredients missing in these ancient texts are tomato and chili which are both colonial entries. Portuguese introduced both tomato and chili in the 16th century. The chili got quickly adopted, the tomato got accepted later and as we see today, there are very few Indian curries without tomato and chili. Now, travelling forward to post independence India, the chicken tikka got grafted into this creamy curry composed of butter and tomatoes, the invention of Butter chicken.

The same chicken tikka in England got grafted into gravy made with local tomato sourcing such as tomato soup, pureed tomato along with cream instead of butter. It is a tangier creamier version with additional sweetness that has since captivated the imagination of Indian diaspora across generations. So were the two I cooked for. These diaspora kids and many like them have grown up on this dish as that one comfort dish to go to since their formative years and the saga continues. So both these dishes have the same ingredients just sourced differently. The one thing that seems to have survived as an ingredient on meat since Ancient India, fenugreek, is added as the final garnish in both curries balancing the sweetness and possibly paying homage inadvertently.

In ancient times, the domesticated fowl was chastised and the jungle fowl was the royal food as the king hunted it. More importantly, the meat of other animals ruled the roost. Now the chicken arrived at everyone's table with this one iconic dish, Chicken tikka masala. It has captured the hearts of all classes and nationalities alike. You rarely see an Indian kitchen not have it on its menu.

At the end of this fantastic journey of this dish, there has been some ugly fights. Two families fighting over who invented butter chicken. Two communities vying over the invention of Chicken tikka masala in England. In fact there is a lawsuit over trademark on butter chicken in the court in Delhi. But the rest of the world has moved on making their own versions not caring about any of this proprietorship. It is loved and cherished across generations all over the world. These fights are petty because the culture always owns it. Even the yogurt marinade is not an ownership issue. Food always finds a way to amalgamate cultures and make the seams disappear.

The palate ultimately unites all.

Coming back to my own cooking of this dish. I had a conundrum. I had dismantled my DIY tandoor because of smoke and maintenance issues. I had to improvise. So, I made a makeshift tandoor oven with my outdoor pizza oven. I got the marinade down but had to solve the smoky aroma problem as the vertical drip is not possible. So I had to have burning charcoal stay on the silver foil the chicken was placed on, so the juice will hit them to get that smoky delicious aroma infusing. It turned out pretty good and I dropped them in the gravy. Now the usual ingredients, crushed tomato, a pureed base of cooked spices and onion, cream. Finally my secret ingredient, tomato ketchup, the Indian kind without corn syrup. Garnish with fenugreek. I inverted the vessel but kept the planks and the constant, fenugreek.

So a spiced meat travelled through thousands of years changing itself, its coating and its vessel and created its own sibling. All along it kept its planks made of wood of different grains and the same fenugreek as the mast.

Then there are people still fighting over the keel.

Notes

Why yogurt works on meat. It tenderizes less through acid than through calcium, which activates the meat’s own calpain enzymes and lets them break down connective tissue from within. The mild acidity helps the proteins unwind and hold water rather than expelling it, which is why yogurt gives juicier meat than a stronger acid like lemon or vinegar. And the coating browns: milk protein and lactose together drive the Maillard reaction on the surface, while the fat carries heat and flavor compounds into it.

Curd in the Mānasollāsa meat preparations. Verses in the 1440s, third Viṃśati, chapter 13: meat, fat and marrow formed into pieces as before, in churned buttermilk with mustard and citron pith; well smoked with hing, the pieces are put into curd; mixed with curd and tamarind, or the essence of pomegranate.

Sources. Suśruta Saṃhitā, Sūtrasthāna 46, in Bhishagratna’s translation, for the śūlya preparations and kandupachita. Mānasollāsa of Someshvara III, third Viṃśati, chapter 13 (Annabhoga), in G.K. Shrigondekar’s edition, Gaekwad’s Oriental Series No. 84. Puranānūru 364 and Porunarāṟṟuppaṭai 103–104, in Vaidehi Herbert’s translations.

The trademark is Daryaganj's, "By the inventors of butter chicken and dal makhani," registered in March 2018 on a proposed-to-be-used basis. Moti Mahal filed suit in the Delhi High Court in January 2024, running to 2,752 pages and seeking damages plus a declaration of sole inventorship, and has since filed a separate plea to cancel the trademark. The case remains unresolved.

I am a curious individual with a scientific temper and rigor, and wine has been something I have always wanted to conquer from a rationalist point of view. I have always rendered the field as a subjective phenomenon and had been very skeptical about its scientific validity. But I have too many acquaintances and friends who I would call reasonable people well engaged in that world. So instead of falling prey to my own cynical surrender, I decided to resist. My friend, a kindred spirit, and I ran an experiment a few years ago where we used a well-tested wine tasting set for beginners with journals, flavor charts, and the rest. We earnestly engaged in the process. We were only able to experience some of the references of wine, such as basic tastes or the feel of the body, but real flavors were hard to grasp, and even harder for us to reach a consensus on.

We persisted despite this cognitive dissonance. We tried every variety possible. We got the same results consistently: a much more simplistic perception of wine, sweet or dry, full or light, balanced on the whole from a taste perspective. We would feel flavors all over the map in a single wine between the two of us. It did one thing to us that we did not expect. We realized that wine is indeed very interesting, and in most cases enjoyable, if some basic agreement was met for each one of us. We realized why it has captivated humans for eons. But we still did not understand how, and what in our own body constitution, all these interesting parts are perceived. We were also astounded by the variation between us, who were very similar-minded people.

So I resorted to my weapon of reason: science. The science of perception in humans is governed by a defined anatomy. Being a physician, I had domain expertise in this. So I decided to take an anatomical journey of wine. First I would dive into the macrocosm of receptors, then dwell in the microcosm of perception. Every time we swirl wine in a glass and take that first sip, we start this anatomical journey.

Even before we venture into anatomy, let us make a brief stop to address a common misconception which is constructed with vision. Swirl a glass and watch the film climb the inside wall, gather, and fall back in slow streaks. Wine experts call these the legs and instruct the drinkers to read them as a sign of richness, quality and conclude it as a wine worth respecting. They are nothing of the sort. The legs are alcohol and only alcohol. As the swirled film sits against the glass, the alcohol in it evaporates faster than the water, and that imbalance drags the liquid up the wall until it beads and slides back down. More alcohol causes more legs. Sugar content changes only how slowly they crawl, not whether they form. The eye can read only one crude fact, the strength of the drink, but it is being coached to see something finer that cannot be seen.

You can prove it in ten seconds without spoiling a drop of your precious wine. Swirl the wine and then cover the mouth of the glass with your hand. Sealing it off from the air makes the evaporation stop, and within moments the legs stop forming. Lift your hand and swirl again. The legs will climb right back. Nothing about the wine has changed. You have only switched the physics on and off.

By doing so, you have watched the quality you were admiring turn out to be weather on the inside of a glass.

The color is the second thing the eye reads and here, the manipulation gets worse. We take a deep inky red as a sign of concentration and seriousness. Some of that is real. The riper fruit and longer, warmer growing season can pack more pigment into the skin and depth of color once served as a rough tell for where and how a wine was grown. But most of what you are seeing now is not that. It is because of the grape variety and the number of days the winemaker chose to leave the juice sitting on the skins. This is a human decision made in a cellar and not a verdict handed down by the sun. And here comes the deceit. The winemakers know that the drinkers’ eyes have been led to believe to read darkness as quality. So they reach for it deliberately by extending skin contact and choosing techniques that deepen the color of the pour. The signal is not merely misread but it is engineered to be misread. The alphabet of the eye which is just three letters wide, is being written in on purpose by someone who knows exactly how the instrument will read it. Climate change, meanwhile, is scrambling what little geographic meaning the color once carried and therefore, even the honest part of the signal is coming loose.

Back to the anatomical route of our first sip, the tongue is the first place it lands and it barely registers. The tongue reads five things, but wine leans on only three of them. Sweetness is perceived when the sugar is left behind by the yeast. Sourness is felt as the acidity a taster means by crisp or bright or racy. The tannins bring in the bitterness. Those three are the working spine of any wine on which my friend and I agreed. The other two tastes barely enter the scene. Saltiness is felt faint at best as wine carries little sodium. Umami is the savory note we borrowed from Japanese and is real but quiet due to a low hum from the amino acids in the grape. Those acids are coaxed up a little in wines that have aged or rested on their lees. That is all the tongue can say, wine speaking in three of the five words it has. My friend and I can agree here because a few letters make it easy for two people to read the same way.

The tongue is a modest apparatus which is roughened by bumps called papillae, and inside three of the four kinds sit the taste buds, each a cluster of fifty to a hundred and fifty receptor cells. These receptor cells are not neurons. They are epithelial cells, skin’s kin, and they die and are replaced every week or two. But the nerve beneath them stays. When a receptor cell is renewed, the standing nerve fiber reconnects to the same type of cell it used as a reader before. The same five letters are handed down intact. The reader changes but the readout does not.

There is a word called minerality. Drinkers reach for it constantly associating it with the taste of wet stone, of slate, of crushed rock, and speak it as though it named something precise on the tongue. It does not. There are no mineral receptors. When the ions in wine, the magnesium, the calcium, the traces of iron, register at all, they are perceived across the machinery as a confused blend of bitter and metallic and faintly salty, dominated usually by bitterness, and much of what we call minerality is not taste at all but smell, or suggestion, or the name itself doing the work. It is a celebrated descriptor floating free of any instrument built to read it. It is the legs and the color moved indoors onto the tongue. A word people trust because everyone else uses it. It is pointing at a sensation the body was never equipped to spell.

The next stop on the route is the palate and here the wine works by touch as much as taste. Tannin does two jobs at once. It registers as bitter, and is also felt as a grip, drying the cheeks as it binds with saliva. That grip rides the touch nerve, the same nerve that reports the weight of the wine, its body, whether it sits light or full in the mouth. My friend and I mostly agreed here too, on body, on grip, on the architecture of a wine. But the agreement already had a crack in it, because touch is not identical between two mouths.

The clearest of these cracks has a name. There is a gene, TAS2R38, that largely sets how loudly a person registers bitterness, and it comes in two common versions. One reads the bitter compound; the other is nearly deaf to it. The split is not rare or exotic. It runs close to down the middle of the population, so of any two people pulled off the street, one may taste a bitterness the other can barely find. The sensitive ones, the so-called supertasters, can register that bitterness far more intensely, and they drift, without ever choosing to, toward sweeter, softer wine, because the dry, tannic, serious reds that a connoisseur calls profound arrive on their tongue closer to an assault. The gene is most of the story, though not quite all of it; a second receptor and the sheer density of taste buds add their own weight. But the shape of the thing is already clear. My friend and I were not lazy readers who failed to agree. We were, in part, two different instruments handed the same glass, each reading a bitterness the other was built to read differently.

And we came to see an old injustice in this. For too long the sweet-wine drinker has been written off as the less sophisticated palate, the beginner who never grew up, by people who mistook an accident of their own receptors for a virtue. But the person reaching for the sweeter glass is often simply obeying a gene, tasting a bitterness the connoisseur cannot feel and declining it, sanely.

To call that unrefined is to call someone’s blood type a failure of taste.

The ignorance was never theirs. Then there are people in the middle who are confused by all this and may choose not to participate at all.

And yet we were close. Fewer alphabets and a stable reading instrument make it near enough to argue productively, near enough to agree on the shape of a wine. The scattering had not started. It started the moment we left the tongue and the palate behind and tried to name what we smelled.

Every other sense can be handed to you in a dictionary. Only smell must be lent as a phrasebook, and the phrasebook is being rewritten while you sleep.

Consider how little the other senses ask of you. Sight begins with three kinds of cone, and every color you will ever meet is a chord struck among those three, so a painter can name a hue and trust it lands the same in your eye as in hers. Hearing is a keyboard strung along the inner ear, low notes at one end and high at the other, an arrangement fixed for a lifetime. That is why a score set down two centuries ago still plays the same.

The tongue shows the trick most plainly. It carries some twenty-five separate detectors for bitterness alone, and quinine and grapefruit and a hundred plant alkaloids each find their own, and every one of those detectors reports back the same single word. Bitter. Twenty-five instruments, one syllable out, and everything about which of them fired is discarded on the way.

That is what a sense does when it works the way we expect it to. It takes a wide world and squeezes it onto a short list, and the list is identical in you and in me because the squeezing is done by anatomy and not by us. Three primaries. Five tastes. A settled handful for touch. That short shared list is what a dictionary is made of, and it is why each of them can publish one and be believed.

Smell does not squeeze. Four hundred detectors sit in the nose and beneath them lies no short list at all, and not for want of looking. Henning drew his prism of odors, Amoore counted out his primaries, and a century of serious people have gone hunting for smell’s five words. There are none. We cannot yet agree on how many dimensions the space even has. Every other sense hands the brain a small tidy code, and smell hands it four hundred numbers and no key.

Smell is built like a fluid, taking the shape of whatever glass the reader holds it in. Here the receptor is not a disposable cell wired to a permanent nerve; the receptor is the neuron, a true nerve cell with an axon that runs straight into the brain. And this entire neuron is what dies and regrows from stem cells at the roof of the nose. This cycle repeats season upon season because it alone stands exposed to the open air and must be replaced as fast as the world burns it away. Each new neuron must find its way and wire itself again. So the sheet that reads a wine tonight is not, cell for cell, the sheet that read one a year ago.

Which means the most cruel part of the subjectivity is not the one between us. It is the one that exists within one person’s life, and I am not referring to the reduced acuity that comes with old age but the one that is happening this month against last month.

Consider what it means that the reading apparatus is replaced on a rolling schedule. The nose you carry to the glass tonight is not, in its cells, the nose you carried a season ago; a fraction of those neurons have died and been regrown since, each newcomer having to find its way and wire itself in from scratch. There is no master copy being faithfully recopied. There is only a reader under perpetual renovation, learning the room again as it goes. And the thing it checks the wine against is no steadier. The remembered violet, the reference you reach for to name what you smell, is not filed away intact and pulled down whole. It is reassembled each time you recall it, rebuilt from the same broken catches, a newer version standing in for the earlier one you can no longer reach. So the act of recognition sets two unstable things against each other: a live reader that was partly rebuilt since the last glass, and a memory that is rebuilt in the very act of remembering. When they seem to match, we call it recognition.

We are matching one reconstruction against another and trusting the fit.

This is why a taster can swear a wine has changed when the bottle is identical. Nothing in the glass moved. The reader did, quietly, between one tasting and the next, and the standard it was measured against moved too. She is not lying, and she has not simply lost her edge with the years. She is, this month, no longer quite the same instrument she was last month, reading against a memory that is no longer quite the same memory.

The wine you could confidently still name is the one you have most likely lost. What follows is my rational interpretation rather than anyone’s finding, so take it as reasoning and not as a result.

Drift runs at a rate, and exposure sets it. Smell a thing often enough and the pattern holds its shape; stop, and the sliding speeds up. Which means the sommelier who tastes every morning is not maintaining her references at all. She is maintaining her instrument. The vocabulary stays warm, the attention stays practiced and the ritual holds. But the glass in front of her tonight is being read against a trace laid down once, at a table, years ago, and nothing since has come back to refresh it. Her Burgundy is safe. She drinks it weekly. The bottle she met a single time, the one she would recite from memory if you asked her, is the one with nothing left holding it in place.

So two tasters of identical training, both working daily, both scrupulous, still carry reference sets decaying at rates set by nothing grander than how often each wine happened to cross their table. They will disagree about the same glass. Neither of them is careless, and neither is lying, and the disagreement was built into them by their internal diaries.

Set that at a table and you see why no tasting was ever a shared reading. Every seat holds a different nose, drawn from a different hand of the same few hundred genes, and every one of those noses is drifting from its own past even as it lifts the glass. The wine is one. The readings are as many as the chairs, and each reading is provisional against the taster’s own yesterday. This is not a failure of manners, or training, or nerve. It is the floor of the thing, laid down in the body, and no amount of expertise will lift it.

And the strangest part is that the nose is not failing. It is doing precisely what it was built to do.

Consider the problem evolution actually had to solve. Not a fixed set of smells to be memorized once, but an open and shifting chemical world, thousands of odors, most of them never yet encountered, changing as the animal moved into new country. A fixed instrument would have been useless the moment the herd migrated or the season turned. So the nose was built the opposite way: an enormous signal matrix, four hundred signals and their vast permutations, so that almost anything could be read; a reader that rebuilds itself on a rolling schedule, retuning to whatever territory it finds itself in, growing more of the receptors it is actually being asked to use; a private instrument, calibrated to one body in one place, learning the particular smells of its own life. Every one of the three flaws is a feature. Too many signals, so that no smell is beyond it. Too unstable to hold still, so that it can always learn the new.

And too private to hand off cleanly, which turns out to be the subtlest design of all. Watch any animal teach its young to hunt. The mother does not, because she cannot, transfer the smell itself. What she passes down is the grammar, this class of scent means prey, that one means danger, this means the water is near, and the infant takes the grammar and builds its own private vocabulary underneath it, a language no other creature will ever read, and one that will go on reconstituting across the whole of its life. This is not a broken handoff. It is the only kind of handoff a plastic instrument permits, and it is enough. The knowledge that matters, the grammar, is heritable and shared. The vocabulary stays private and provisional, because it was always going to be rebuilt anyway. Nature never once tried to standardize the word. It passed the grammar and trusted the pupil to write the rest.

We are the ones who broke that arrangement. We took this private, plastic, brilliantly adaptive instrument and asked it to do the one thing evolution never asked of it: to deliver a fixed and shared vocabulary, the same word in your nose as in mine, the same this year as last. We built rankings on it and prices along with a whole architecture of authority, resting on the assumption that a wine smells the same to the trained and the untrained, the same across a table, the same across a life. It does not, and it cannot. Even a lioness teaching her cubs knew better than to try to hand over the word. We conscripted a tool of personal survival into a system of collective judgment, demanded a precision of transmission that no nose has ever used or needed, and then called the people whose noses would not conform unrefined.

The success is the nose’s, and it is total. The failure is ours, and we built it, through wine.

There remains one question, and it is the uncomfortable one. If the nose has always worked this way, and the mismatch has always been there, why did a culture as obsessive as wine’s never see it?

For most of the three thousand years, the answer is simple and forgivable: nobody could have. The knowledge did not exist. The anatomy hides the trick, flavor feels like it happens on the tongue even though it is built in the nose, so every generation inherited the same honest error and passed it down intact. You cannot blame a medieval monk, or a nineteenth-century Bordeaux committee, for not knowing about the olfactory receptor genes that would not be identified until 1991, or the bitter-taste gene that would wait until 2003, or a regenerating neuron that would not be understood until our own lifetimes. The confusion was real, and it was innocent.

But the signs have been arriving for a while now, and they have not been subtle. In 2001 a Bordeaux laboratory dyed a white wine red and watched a room of oenology students describe it, almost to a person, in the language of a red, their eyes silently overruling their noses. The same year, the same wine poured under a grand cru label and a table wine label drew opposite verdicts from the same tasters, complex and balanced from one bottle, faulty and simple from the other, the wine identical in both. Texas A&M once poured a single Texas wine into three glasses labeled France, California and Texas, and watched the tasters rank them in exactly that order. The evidence that the verdict lives in the label and the color and the expectation, not in the glass, was demonstrated in plain view.

And now the mechanism itself is known. We can name the receptors, count them, watch the neurons die and regrow, read the gene that splits the room on bitterness. The excuse of ignorance has expired. What was once an honest confusion is now a convenient one, kept alive not because we cannot know better but because an extensive scaffold, the rankings, the prices, the authority, the flattery of the trained palate, has too much resting on our not looking.

And here is what the fog costs, beyond justice. A culture that insists on one correct palate is a culture busy turning away its own future. It tells the sweet-wine drinker she is a beginner who never grew up, tells the uncertain middle they do not belong at the table, and stakes the whole enterprise on a narrow, aging class of the properly trained, a cohort that is thinning year by year and finding no one behind it. That is not only unkind. It is a business slowly refusing its own customers.

The science offers the way out, and it happens to be the honest way. If the palate is truly personal, written partly in the genes and redrawn every season, then there is no failed drinker, only different readers of the same glass. The one reaching for the sweeter wine is not unrefined; she is obeying a receptor she did not choose, and her preference is as legitimate as any expert’s. The confused middle were never disqualified; they were only told a lie that pushed them out. Clear that fog and the room grows. Every palate the old hierarchy shamed into leaving is a palate that could be welcomed back, not by lowering the art but by telling the truth about it. A category that honors many kinds of drinker is a wider and sturdier thing than one propped on a single vanishing ideal. Diversity is not a threat to wine. It is how wine survives.

Because in the end a wine and its drinker are in a kind of relationship, and relationships last only on honest terms. Tell someone their pleasure is wrong and they will eventually leave. Tell them the truth, that the glass is real, that their reading of it is real, that no two of us will ever get the flavor quite the same and that this is the wonder of it and not the flaw. The nose was built never to hold still, to keep taking in the new and revising what it knew. That was its genius. And a culture is no different. We cannot afford to stay frozen around a comfortable old error once the science has moved, any more than the nose could afford to stop learning the smells of a changing world. The reading was not designed to be stable. Neither, in the end, are we.

A note on sources

The two 2001 experiments come from the Faculty of Oenology at the University of Bordeaux. Morrot, Brochet and Dubourdieu, The Color of Odors, published in Brain and Language, is the white wine dyed red and described by fifty-four tasters in the vocabulary of a red. Brochet and Dubourdieu’s companion paper, on wine descriptive language and the cognitive specificity of the chemical senses, is the same Bordeaux poured under a grand cru label and a table wine label and drawing opposite verdicts.

The three-label demonstration at Texas A&M is described in Wes Marshall’s Liquid Assets column in the Austin Chronicle. The same effect has also been tested formally: Guidry, Babin, Graziano and Schneider, Pride and Prejudice in the Evaluation of Wine?, International Journal of Wine Business Research 21(4), 2009, in which 257 participants rated the same wine under a French and a Texas label. Their finding carries a wrinkle worth naming, which is that identification with Texas could override the prestige of the French label.

The bitterness gene is TAS2R38, identified in 2003 and the best-studied of the taste receptor variants. The olfactory receptor gene family was described by Linda Buck and Richard Axel in 1991, work that took the Nobel Prize in 2004.

For the failed search for smell’s primaries: Hans Henning’s Der Geruch of 1916 proposed the odor prism, and John Amoore’s stereochemical theory of the 1960s proposed a set of primary odors. Neither survived. The field still has no agreed count of the dimensions of olfactory space.

On individual variation, Mainland and colleagues, writing in Nature Neuroscience in 2014, found that any two people differ functionally at roughly thirty percent of their olfactory receptor alleles. On instability within one life, Schoonover, Fink and Axel, in Nature in 2021, tracked piriform cortex representations of the same odor and found a classifier trained on early responses performing at chance after thirty-two days, with repeated exposure slowing the drift without arresting it.

This is the overture. The series that follows takes wine apart from the ground up, from the soil to the glass.

I. A SYSTEM IS NEVER CULTURALLY NEUTRAL

Every evaluation system encodes the culture that built it. This is not a flaw. It is a structural inevitability. The question is not whether a rating system carries cultural assumptions. Every one does, necessarily, because the people who designed it, the vocabulary they used, the culinary tradition they drew on, and the excellence they were trying to recognize were all culturally specific. The question is whether the system acknowledges this, and has the humility built in that the acknowledgment demands.

Michelin does not. Its authority rests, explicitly and consistently, on the claim that its criteria represent a universal standard of culinary excellence, one that can be applied with equal validity to a kaiseki restaurant in Kyoto, a biryani house in Hyderabad, a mole kitchen in Oaxaca, and a bistro in Lyon. This claim is the one that cannot survive honest scrutiny. None has been done by Michelin, and not often enough by the food culture that has accepted its authority.

The cultural bias embedded in Michelin is not subtle. It is structural, historical, and traceable to the system’s origins. Understanding it requires going back to where the system began.

Where the Guide Came From

The Michelin Guide was created in 1900 by André and Édouard Michelin, two French industrialists, to encourage French motorists to drive more, thereby wearing out more tires and generating more revenue for the Michelin tire company. The first guides were given away free. They listed petrol stations, mechanics, and places to eat along French roads. The restaurant ratings were, in their origin, a tire company’s marketing device.

The criteria that developed over the following decades were developed by French inspectors, within French culinary assumptions, at the precise historical moment when French haute cuisine claimed, and was widely granted, the status of the world’s definitive culinary tradition. This was not a neutral historical context. It was the apex of French cultural dominance in Europe, expressed in food as in everything else: the assumption that what France did was the standard against which everything else was measured.

That assumption is embedded in the system’s DNA. It has never been formally removed. It has been papered over by geographical expansion and the occasional gesture toward local culinary traditions, but the foundational vocabulary, the training of inspectors, and the implicit hierarchy of techniques and ingredients that underlie the three pillars remain essentially unchanged from a framework developed in early 20th century France.

Michelin did not invent a universal standard of culinary excellence. It invented a French standard, and then exported it to the world as though the two things were the same.

II. THE SIX EMBEDDED BIASES

The cultural assumptions in Michelin’s system are not confined to a single criterion. They run through every pillar, and through the unstated assumptions that sit beneath them. Six are worth naming specifically, because each one operates against a specific set of non-European culinary traditions.

1. The Sauce Bias

French classical cuisine is organized around the mother sauces: complex reductions built over time through technically demanding processes that demonstrate depth of skill. The sauce, in French culinary logic, is where a cook’s mastery lives. Michelin’s second pillar, mastery of flavor and cooking techniques, was built to recognize and reward this mastery.

Many of the world’s greatest culinary traditions produce no sauce in this sense at all. Indian cuisine’s flavor depth lives in the spice application, in the precise sequencing of aromatics in hot oil, in the bloom of mustard seeds and curry leaf, in the slow reduction of onion to a specific sweetness that takes decades to learn to recognize and achieve. Japanese cuisine’s depth lives in fermentation, in dashi made with the restraint that makes everything taste like itself only more so, in the precise temperature control of the broth that lands subtly on the palate. West African cuisine’s depth lives in the layered stewing of palm oil and fermented locust bean and crayfish powder and scotch bonnet, a flavor architecture of extraordinary complexity that produces no sauce a French-trained inspector would recognize as a demonstration of technique.

These are not simpler traditions. They are traditions organized around different principles, principles that Michelin’s vocabulary cannot evaluate because it was not designed with them in mind.

2. The Ingredient Hierarchy

Michelin’s first pillar, quality of ingredients, is evaluated against an implicit hierarchy in which certain ingredients signal culinary seriousness by their presence. Truffles. Foie gras. Dry-aged beef of specific provenance. Specific seafood from specific waters. Single-estate anything. These are the ingredients that read, to a European-trained inspector, as statements of quality.

A perfectly sourced, perfectly prepared rasam, built from tamarind of precise acidity, tomatoes at exact ripeness, black pepper freshly cracked, the correct variety of toor dal simmered to the right consistency, curry leaf from a specific tree, is a demonstration of ingredient intelligence of the highest order. The ingredients are not on the prestige list. The intelligence is invisible to criteria that did not develop the vocabulary to see it.

The same applies to the mole negro cook sourcing the precise dried chilli varieties — pasilla negro, mulato, chihuacle — from specific regions where the soil and altitude produce the flavor profile the dish requires. Or the injera maker selecting teff from the Ethiopian highlands where the grain’s iron and calcium content, shaped by altitude and soil, is part of the dish’s nutritional intelligence. These are ingredient decisions of extraordinary sophistication. They register as peasant food.

Many a cuisine has to dress French to climb the prestige ladder with dishes such as truffle carpaccio or foie gras samosa, even make their restaurant sound French.

3. The Refinement Bias

French haute cuisine evolved as court food. It was designed to demonstrate that an enormous amount of skilled labor had been concentrated on a small, perfect, exquisitely presented portion. Caloric density is low but labor density is high. This is a specific aesthetic arising from a specific social context which was aristocratic display, the demonstration of wealth through the conspicuous consumption of skilled time. The brutal discipline of the brigade inherited from the French kitchens made the labor of consistency transmissible.

Many of the world’s greatest dishes have the opposite logic: maximum nourishment, accessible technique, communal scale. A perfect bowl of pho that has simmered for eighteen hours is not less refined than a French consommé. The fermentation intelligence in a perfectly made kimchi is not less sophisticated than the reduction intelligence in a classical demi-glace. But Michelin’s criteria encode a specific aesthetic of refinement: the tiny portion, the architectural presentation, the visible labor. That aesthetic systematically misreads dishes that express their intelligence in different directions.

4. The Separation Bias

French classical cuisine separates and sequences. Courses arrive independently. Flavors are isolated, presented in controlled progression, experienced one at a time in a specific order designed by the chef. This is not merely a service convention. It is a philosophy of how food should communicate, derived from a specific cultural aesthetic about clarity, control, and the chef’s authority over the dining experience.

Many of the world’s most sophisticated food traditions are communal and simultaneous. South Indian dining, where dal, sambar, rasam, two or three vegetable preparations, pickle, papad, and rice arrive together on a banana leaf, each element in conversation with every other, the diner constructing their own experience through the mixing and proportion that the tradition teaches, is not a lesser form of service. It is a different philosophy of what eating is for: not the sequential revelation of a chef’s vision, but the participatory construction of nourishment within a tradition that trusts the diner’s intelligence.

Ethiopian injera, the shared fermented flatbread that serves simultaneously as plate, utensil, and food, with wats and tibs and gomen arranged across its surface for communal sharing, encodes a philosophy of abundance and togetherness that is as sophisticated as anything in French service. Mezze, dim sum, thali, banchan: these are not approximations of the correct way to eat. They are complete philosophies of hospitality that the Michelin service model treats, implicitly, as lesser formats.

5. The Venue Bias

Michelin inspectors eat in rooms. The physical environment, while technically absent from the published criteria, shapes every aspect of the encounter. What the inspector expects to find, what they are prepared to evaluate seriously, what reads as an appropriate context for culinary seriousness: all of this is conditioned by the room.

The Bib Gourmand category was introduced in 1997, 97 years into the guide’s existence, as Michelin’s belated acknowledgment that food served in informal environments could be excellent. In practice, it became the category where non-European street food traditions were placed when they could no longer be ignored. Good food, but not star food. Worth a detour, but not a journey. The message embedded in that classification is legible to everyone in the food cultures it describes: your best is our second tier.

A hawker stall in Singapore serving char kway teow refined by a single cook over forty years of daily practice, every wok breath timed, every combination of ingredients calibrated through tens of thousands of repetitions, is doing something that no French-trained inspector was initially equipped to evaluate as culinary excellence. When Michelin eventually did award stars to hawker stalls in Singapore, it was treated as a radical departure. The radicalism reveals the default.

6. The Documentation Bias

French culinary tradition is extensively codified. Auguste Escoffier’s Le Guide Culinaire, published in 1903, systematized French classical cuisine into a written canon that gave inspectors a reference point: there is a correct beurre blanc, a correct consommé, a correct béarnaise, and the inspector’s training includes knowing what correct looks like.

Many of the world’s greatest culinary traditions are transmitted orally, through apprenticeship and embodied practice: the hand that knows the right weight of dough, the nose that knows when the fermentation is ready, the ear that knows the sound of a properly tempered mustard seed in hot oil. This knowledge is not less precise than Escoffier’s. It is more precise. It has been refined through more repetitions than any written canon could encode. But it produces no document against which an inspector can calibrate their evaluation. It is, by the standards of a system that emerged from a documented tradition, invisible.

The grandmother who has made rasam every day for fifty years knows something about that dish that no culinary school has ever taught and no guide has ever rated. That knowledge is not less real for being uncodified. It is more real. It has passed the only test that matters across the only timescale that counts.

III. THE SELF-AWARENESS MICHELIN HAS NOT SHOWN

To its credit, Michelin has been pushed into some degree of engagement with non-European culinary traditions by market pressure, by persistent criticism, and by the commercial impossibility of ignoring the culinary sophistication of Japan, Thailand, Singapore, and Hong Kong. The Bangkok guide was a genuine departure. The Singapore hawker star was a significant gesture.

But departure within a system does not change the foundational assumptions of the system. It creates exceptions that prove the rule. When Michelin awards a star to a hawker stall, it is not changing its definition of culinary excellence. It is declaring that this particular hawker stall has met a definition the stall’s cook had no part in setting. The authority still flows from the institution. The validation still requires the institution’s approval. The framework has not been revised. It has been stretched, temporarily, to accommodate a case it was not designed for.

The test of genuine cultural competency in a rating system is not whether it can occasionally recognize excellence outside its native tradition. It is whether the criteria themselves have been developed with enough cultural humility to recognize that they represent one tradition’s understanding of excellence among many, and that other traditions have developed their own understandings, equally rigorous, equally sophisticated, and not reducible to the vocabulary of the tradition doing the evaluating.

Michelin has not passed this test. Its criteria are presented as universal. Its inspectors are trained as though the culinary knowledge they are evaluating is a single body with a single vocabulary. Its authority is claimed as though the standard it applies belongs to no particular culture. None of these things are true. And the food cultures that have been evaluated against these untrue premises, measured, found adequate or inadequate, starred or unstarred or uncovered, have paid the cost of that untruth in the currency that matters most in food: whether their traditions are seen, recognized, and treated as the equal of any other.

IV. WHAT CULTURAL COMPETENCY WOULD ACTUALLY REQUIRE

The answer to Michelin’s cultural competency problem begins from a structural observation: no single evaluator, trained in any single tradition, can produce a reliable verdict on a food culture they did not grow up inside.

This is not relativism. It is a claim about where food intelligence actually resides. The intelligence required to evaluate a rasam, to know whether this one is better than that one, whether the tamarind is the right acidity, whether the curry leaf has been added at the moment that extracts its full aromatic value, lives in the community that has been eating rasam for generations. It cannot be acquired through training in a different tradition and deployed at the table through an act of professional will.

A genuinely culturally competent evaluation framework would define sensory pleasure as the pleasure experienced by the people for whom the food was made, not the pleasure registered by someone trained to evaluate a different tradition. It would define ritual embedding by the community the food serves, not by the service conventions of European dining. It would define ingredient quality by what the tradition in question recognizes as quality, not by the prestige hierarchy of a system built in France in 1900.

The knowledge required to evaluate any tradition of food accurately already exists, in the communities that built those traditions, refined them across generations, and embedded them in the occasions that make a culture legible to itself. What has always been missing is not the knowledge. It is the mechanism to aggregate it.

The Aggregation Problem

The most culturally competent evaluation of any food is the evaluation of the people it was made for. A thousand evaluations of a moqueca by people who have been eating it their whole lives produce something no single inspector can produce alone, however open-minded, however well-traveled. The aggregate knows which element of this specific preparation is slightly off relative to the tradition’s standard.

This is not a romantic claim about the wisdom of crowds. It is a structural observation about signal quality. The aggregate of the people for whom a food was made contains more accurate information about whether that food is good than any individual, including any individual expert in a different tradition, can contain alone.

Michelin’s system cannot produce this signal. It is not designed to. A single inspector, trained in one tradition, visiting once, against criteria that tradition produced, generates a different kind of information: information about whether this restaurant meets that inspector’s expectations. The information is real. It is simply not the same as whether this food is good by the measure of the people it was made for.

What we have as aggregation is corrupted with a lot of noise such as cultural naivety, tourist trap, performance, and so forth. They have solved the volume problem but have not addressed the calibration problem.

The most honest evaluation of any food is the evaluation of the people it was made for. That evaluation has always existed. It has never been systematically collected. The system that would collect it, at scale, across every tradition, from every community that has something to say about what it eats, does not yet exist.

Michelin speaks for the inspector. The system that would actually work speaks for the culture. These are not the same thing. Building the second is the work that remains.

The people who sold tires built this. The culture that knows the food should build the one that serves all.

The dishes left behind

Mole negro (Oaxacan black mole) is the rich, complex traditional mole of Oaxaca, Mexico with its jet-black color, velvety texture, and intricate balance of savory, sweet, smoky, and bitter notes, it often requires over 30 ingredients and days of preparation. Every home makes its own version with the different permutations and the best will be lost if we are just stuck with what Michelin will hand you.

Rasam is a light, spicy, and tangy ancient South Indian soup traditionally brewed with tamarind juice, tomatoes, and a fragrant mix of freshly crushed spices. It is a digestif, a soup and a delicacy all put together.

Injera is a sour, fermented flatbread with a unique spongy texture made out of teff flour that forms the culinary foundation of Ethiopian and Eritrean cuisine. It acts simultaneously as a plate, a lining, and the primary utensil used to scoop up stews and vegetable dishes. It is fermented with ersho, the tangy liquid saved from the previous batch, often inherited from the previous generation. So no one injera will be the same.

Moqueca Baiana is the iconic Afro-Brazilian seafood stew known for its rich, velvety broth, bold golden-orange hue, and tender seafood, and aromatic layers of peppers, tomatoes, and herbs made with palm oil. It represents a seamless fusion of West African and Indigenous Brazilian culinary traditions. Michelin is in Rio but will never see the real version. Coverage does not mean reach.

The Michelin Guide's founding in 1900 by André and Édouard Michelin, and its origin as a tire company's marketing device distributed free to French motorists, are the company's own account of its history. Auguste Escoffier's Le Guide Culinaire was published in 1903. The Bib Gourmand was introduced in 1997, ninety-seven years after the first guide. The first stars awarded to Singapore hawker stalls came in 2016, and the first Bangkok guide followed in 2017. The three pillars, and the criteria described here, are as Michelin publishes them.

I have made many a career decision by moving through uncertainty until a clear path appears. But on that Friday evening, after a long week, a friend’s dinner invitation left me stranded in a wine store, caught in a fog I could not find my way through.

The invitation came with a rule, well-meant and clearly stated: do not bring just any wine, nothing off the bottom shelf, nothing under a certain price, and if you are not sure, bring nothing, we would honestly rather you did. A little blunt, I thought. But it’s their table, their call. And it should have been the easiest errand in the world because the fix looked obvious: stay above the floor they had set, and you should be safe. So, I stood in the aisle with my wife beside me, hunting for the bottle that cleared it, and could not find the place where below became above. I have always decided on wine at the table. Every label promised things I had no way to check. The price told one story and the words on the back told another and the rule I was trying to satisfy assumed a kind of seeing I did not have. We narrowed nothing. We left with nothing. We drove to their house with empty hands, which by the terms of their own rule, was a permitted answer, and somehow the permitted answer felt like the most exposed one of all. What was their bottom shelf and was price a reliable way to get to the top floor?

I had wanted to understand wine for a long time, not to perform it, but to truly grasp it, the way you want to understand anything intelligent people treat as obviously meaningful while you stand at the edge of the glass, nodding and hoping the nod is enough. For years, it sat on my list of things to learn. That aisle was where the bill came due.

The scientific rigor in me and the skeptic that I was did not make it easy. And more than once I have been tempted to settle it the cheap way, to decide the whole thing is theater and walk off lighter. The exit was always right there. But something kept stopping me, and it was not manners. It was the people. The ones who believe in this are some of the most reasonable people I know, and reasonable people do not pour decades of attention and real money into nothing. And underneath them, the longer fact: culture does not keep things alive out of sentiment. Whatever survives this long survives because it is doing some kind of work, even when the people doing the work cannot tell you what the work is. So, the contempt never closed the deal. There was something in the glass I had not found, and walking away would only have meant I stopped looking before I found it.

So I went and took it apart instead. And what I got sucked into was an entire journey within myself: a line the wine travels from the moment it is a molecule to the moment I pronounce a verdict over it, and a string of stations along the way, each one adding something the station before it did not have. I had been standing on a platform my whole life mistaking it for the ground. The first station, the one nearest the front, is the one that should have shown me the rule itself was the problem, not me, not the wine.

I board the first station swirling, sipping, and holding a mouthful with the glass still in my hand, thinking the tongue is where the wine is being read. The tongue does not and cannot taste the flavors of wine. The tongue is a blunt instrument, and this is not a figure of speech, it is anatomy. Sweet, sour, salt, bitter, and the savory one we had no word for until we borrowed umami. That is the working range of the human tongue, five notes, maybe a sixth if you are generous about fat. It does its part at this station, but is small and gives a structure for the journey. It catches the sourness (the acidity a taster means by crisp, or bright, or racy) and that spine is real, on the tongue. It catches sweetness when the yeast leaves sugar behind. It catches a low hum of the savory in wines that have aged or rested on their lees. And that is nearly the whole of it. But those few notes have to agree, and when they do, that agreement is the spine holding, the balance the rest of the journey is built on. Everything else, every word that has ever made wine sound like poetry, is happening somewhere other than the tongue. Blackberry. Graphite. Saddle leather. Crushed violets. Wet stone. None of it is taste. It is smell: aroma climbing the back of the throat into the nose while you hand the credit to the tongue, laid over memory, laid over a vocabulary that arrives a half-second behind the sensation and tells it what it was. The sommelier swirling and pronouncing is not reading her tongue. She is reading her nose, and her nose is wearing the tongue’s name, because the tongue is the part that touches the wine and gets billed for it.

I knew this in the store. That is the part I keep returning to. I am not usually frightened by a sophistication I do not possess. It only makes me curious. I had read enough to know that tongue is a courtesy title, that the friend whose table we were driving toward could not literally taste the bottle I failed to bring any better than I could. And the feeling of something uncertain sat exactly where it had been, untouched by everything I knew, which was the first sign that it had never been about the wine. Because once you stop being charmed by the misnomer, it opens onto something larger. If we are not tasting these things, if violet and leather and stone are smell and memory and a handed-down word, then the whole edifice, the rankings and the prices and the reverent silences, stands on a sensation we have been mislabeling since time immemorial. A gap that old, that load-bearing, that universally agreed upon, is not a mistake. Nobody sustains a mistake for three thousand years unless it becomes the structure itself. Everybody has been building in that gap. The question that kept me standing there was not why we cannot taste it. It was what we put in the space where the taste was supposed to be.

As I traverse to the next station, I find a different landscape but the same ground. Here the wine does play on my taste notes a little through the tannins that wear two hats, both touch and taste. The tannins taste bitter here, and teaming with saliva, act on the touch nerve, and I feel friction. That is the dry grip the red leaves in my cheeks. The same nerve also feels the weight and viscosity of the wine. Now you can notice that the ground here is the same: the touch nerve. But the passengers are different: grip (astringency, friction) and body (weight, viscosity). A wine can be light and grippy, or full and smooth. The way the passengers behave can vary with different mouths. There is a gene, and you can have it tested, that decides how loudly you taste bitterness. The people who carry the sensing version taste it fiercely such as the bitter edge in coffee, in greens, and tend, without ever deciding to, to drift toward sweeter wine, because the dry, tannic, “serious” reds arrive on their tongue as something closer to an assault. We call them supertasters, which sounds like a gift and is closer to a curse. So, the sweet wine drinker is not an inferior drinker, which is like saying someone’s mouth is wrong. Likewise, the grip can be stronger in certain mouths because of the density of the sensory touch hardware. So, the same passenger can behave differently on the same ground based on which carriage they are in.

Which means the dry red that the connoisseur calls profound is, for the person across the table, a literal mouthful of bitterness that their body is built to reject. The whole construct of “developing your palate,” and of being taught to prefer the dry over the sweet, may be, for a significant fraction of people, a quiet instruction to override their own receptors and call it sophistication. The sweet-wine drinkers who are labeled deficient simply do not want to taste something they were born unable to tolerate. There is no single wine in the glass to be right or wrong about. The floor itself is plural, dealt to each of us before we ever chose anything, and my uncertainty in that aisle that the fifteen-dollar bottle was fine and the fear was foolish was resting, the whole time, on a receptor I did not pick.

Now, after I have tasted and felt the wine, I move to reconciling it in my mind. That is the next station down the line, and the wine cannot reach meaning without passing through it.

Smell is the loneliest of the senses. You cannot point at an aroma the way you point at a color; you cannot hand it to someone and ask if they see what you see. What you catch when you lower your nose into a glass is faint, and fast, and gone, and private. And left alone, my violet and your violet have no way of meeting. So culture built the thing that lets them meet. The vocabulary, the reference notes, the training that teaches a roomful of strangers to land on the same word for the same fleeting smell: all of it is machinery for turning a private flicker into something two people can say to each other. This is not snobbery. This is the genuine achievement underneath the snobbery, the part that is pure infrastructure: the means by which a sensation that happens alone, behind your own face, becomes a language. When a sommelier teaches you to find the violet, she is not lying to you. She is handing you the key that makes your nose legible to other noses. That is a cognitive contract, and I want to be fair to it before I say the next thing.

The next thing is that there is nothing solid underneath the agreement to anchor it to.

I had assumed, the way everyone assumes, that the wine words sit on top of some firm thing in the glass: that “violet” is the steady bedrock and the language is the flag we plant in it. It is not steady. Ionone, the compound that smells of violet, has a strange property: it saturates the receptor almost the instant it touches, so the smell does not hold, it pulses: you catch it, it vanishes, it comes back. And the violet you are matching it against, the one in your memory, was itself assembled from those same broken catches, edited by your brain into a steadiness the actual flower never offered. So when you say a wine smells of violet, two unsteady things are happening at once. A flicker in the glass and a flicker in your memory, and the moment they fall briefly into phase, your mind extrapolates a feeble recurrence to fill the gap, the same way it fills your blind spot and hands you one continuous scene. It gives you a violet that sits there, solid, certain, a steadiness that exists nowhere in the world and only in the act of perceiving it. We are not matching a sensation to a fixed note. We are tuning one trembling thing against another and calling the resonance truth.

There is a word for what wine does to you after enough of this, and I reached for it one night and found the language had misplaced it. Ebriated. There is no such word in the English anyone actually speaks; the drunkenness only ever arrives already prefixed, inebriated, the way the violet only ever arrives already steadied. I had coined a word that did the exact thing I was trying to describe, an approximation that creates a reality, a frame thrown across a gap, summoning a meaning the substrate never quite supplied. And you understood it anyway. That is the whole mechanism, performed in a single broken word: the steadiness is manufactured, and the manufacturing is not a flaw in perception. It is perception. There was never a clean version underneath to compare it to.

This is the point where most people decide wine is a con, and I understand the temptation, because I have felt it. But a con needs your ignorance to work. Show someone the card up your sleeve and the trick dies. The strange thing about wine is that it survives the exposure. You can know, fully and in advance, that naming the violet will make you more likely to smell it, and you will still smell it. The swirl does not stop releasing the aroma because you have read the chemistry. The misdirection is not pointing you away from the truth, the way a magician pulls your eye off his hand. It is pointing your attention at something it would otherwise have walked straight past. And attention, it turns out, is upstream of perception: you cannot smell what you are not attending to, and the back-of-the-throat aroma is faint enough and brief enough that without the word to hold your nose still, you would lose it entirely. The performance is how the palate thinks. There is a difference between a lie and a frame: a lie tells you something false, a frame tells you where to look. Most of what gets mistaken for wine’s dishonesty is just frames, frames that point at something real. A sunset does the same thing. You know it is just dust and pollution scattering the short wavelengths out, leaving the red, but the magic does not drain out of the sunset.

And then there are the frames that point at nothing in the glass at all, and those are a different animal, and they are what my friend’s rule was made of. They live at the last station on the line, the one with the longest platform and the blinding lights, where the wine arrives already carrying its molecule and its shared language and gets handed one more thing: a verdict about the person holding it.

Because the agreement, once it exists, gets bolted to things that have nothing to do with smell. A clean signal comes off the calibration, this is what this wine is, and then someone welds it to a slope, a name, a year, a border, a price. The monks who first walked the hillsides of Burgundy, tasting one parcel against the next and walling off the better ground, were welding in good faith; the place really was in the glass, and they were its witnesses. But the weld migrates and it curdles. When a nineteenth-century committee ranked the great estates of Bordeaux for an exhibition, they did not taste them into order: they sorted them by the price the market was already paying and froze that into a hierarchy of quality that has barely moved since. A price list, fossilized, wearing the mask of the tongue. That is the same operation as the monk’s and the opposite of it, and from the outside the two look identical.

What I came to see is that there are really two lines running through the same country, and the trouble starts when you cannot tell which one you are on. One line carries the wine up through meaning: molecule, then language, then status. The other carries it up through quality: clean, then good, then genuinely rare. A passenger boards believing he is climbing the quality line and is quietly switched onto the status line somewhere in the dark, and pays the fare for one trip convinced he took the other. The weld is that switch in the track. And once you have seen it thrown, you start to catch it everywhere.

I have watched the false weld do its work at a dinner table, the violence of it disguised as taste. A good whisky, enjoyed, admired, half-finished. And then someone learns where it was made, somewhere unfashionable, somewhere the story says good whisky does not come from, and the glass is poured down the sink. Nothing in the liquid changed. The geography walked into the room and traveled backward through time and spoiled a thing the man had been enjoying a minute earlier. He did not taste a fault. He failed a map. This is not anecdotal, nor a moral indictment. There have been studies done to verify it. In a 2001 experiment at the University of Bordeaux, the same midrange Bordeaux, poured from one bottle labeled grand cru and another labeled table wine, was scored “excellent, complex, rounded, balanced” from the first and “weak, short, flat, simple, faulty” from the second. In another study that year, a white wine dyed red with an odorless dye was described, by fifty-four tasters, in the language of a red. And in a 2008 study, subjects who tasted wines in a functional MRI scanner rated the identical wine as more pleasant when it was framed at a higher price. The scan showed the brain’s pleasure response rise with the price, indicating that the cost altered the experience, not just the rating.

This train journey I experienced started with balance and harmony. It then transitioned into body and smoothness. Along the way I found a way to be inclusive of how differently mouths are built, because none of us are made equal there. Then the journey got interesting at the flavors, where the sommelier is a guide who hands you the olfactory keys, not a verdict, because flavor, in its truest sense, is deeply personal. That is where the wine gets interesting. So far, so good. But then the deceit and the pollution start. This is where people get alienated, or seduced onto an unsustainable path. It is like any relationship: fakery and deceit do not make it last. Fewer people are drinking wine. The losses at the boomer age who have swallowed the status quo are not being replaced and the people who would replace them are not being met with an open door. The wine industry, though, is doubling down on snobbery, blinded by a narrow vision.

What should be the weld on the track? Meritocracy and transparency. Like Europe is now mandating, wineries should list their ingredients and additives. Beyond that, they should be transparent about their farming certification, their yield, their vinification, where the fruit was sourced, and how the wine was stored. This gives a buyer a clear barometer to gauge earned quality and to see through unearned price.

And finally we arrive at the engine that drives it all. The essential ingredient that seduced our ancestors in the first place. It is the alcohol that is detected as warmth in our mouth, and eventually the one thing you cannot fake: drunkenness. In the ritual of this entire journey, the most honest thing is alcohol.

Now, coming back full circle, was my friend wrong in creating that criteria? Yes and no. He was wrong on the count of price. The chance of failure is much higher when someone uses price as a metric. The floor is not likely to be defined by price. Your chance of success is far better using the meritocratic barometer. Where he was right was the bottom shelf, because where a wine sits is a real tell. But even that only gets you to quality. What he did not mention is the flavor of wine. As we discussed through my journey, it is very personal. So know what he has emptied in the past, or the wine producer he prefers, before you buy wine for him, or for anyone for that matter. Because you cannot go by what you perceive or what the label says.

Sources

Morrot, G., Brochet, F., & Dubourdieu, D. (2001). The color of odors. Brain and Language, 79(2), 309–320. (The white wine dyed red, described by 54 tasters in the vocabulary of a red.)

Brochet, F., & Dubourdieu, D. (2001). Wine descriptive language supports cognitive specificity of chemical senses. University of Bordeaux, Faculty of Oenology. (The same wine described oppositely under grand cru and table-wine labels.)

Plassmann, H., O’Doherty, J., Shiv, B., & Rangel, A. (2008). Marketing actions can modulate neural representations of experienced pleasantness. Proceedings of the National Academy of Sciences, 105(3), 1050–1054.

Wine consumption figures: Wine Market Council, 2025 U.S. Wine Consumer Benchmark Segmentation Survey (share of legal-age adults drinking wine every two to three months fell from 34% to 29%; roughly 76 million consumers, about nine million fewer than 2023). U.S. wine sales down approximately 6% in 2024 (SipSource).

I’ll keep its modern name to myself for a while, if I may. You may have discerned it by now or you may have not. It will be the better for us to meet it as its makers first knew it, and then follow it forward.

The oldest trace of this dish as a word surfaces in a Kannada text around 920 CE, iḍḍalige, a steamed cake of black gram, with no rice in it at all. Two centuries later the same cake reappears in one of the strangest cookbooks ever written called the Mānasollāsa, an encyclopedia of royal pleasures compiled around 1130 by a Chalukya king named Someśvara III, who ruled the western Deccan. His book pauses between chapters on elephants and palaces and the correct temperature for bathwater, to set down how his kitchen made its cakes. He gives three recipes that have bearing on our story, and they are worth reading in his own words.

The steamed cake — iḍḍarikā.

māṣān vidalya cūrṇāni kṣālitāni vibuṣya ca |
vastrāvapūtaṃ santaptaṃ kāñjikenāvimiśrayet ||
amlatvaṃ gataṃ piṣṭaṃ golakīkṛtya mātrayā |
śālipiṣṭena saṃveṣṭya piṣṭam anyat tu vāsasā ||
svinnaṃ caiv’ eddarikā sūdāryaiḥ parikīrtitā |
hiṅgu-jīraka-sampannā dhānyāka-saṃyutā hi sā ||

Split the black gram, grind it, wash and husk it, and strain it through cloth; then mix it with warm sour gruel. When the paste has turned sour, form it into small, even balls; coat them thinly in rice flour, or else wrap them in a wet cloth. Steamed, this is what the master cooks call iḍḍarikā — seasoned with asafoetida, cumin, and coriander.

The surprise lands as soon as you read it. There is no rice in the cake. The body is black gram: washed, ground, soured, and steamed. Rice appears only on the outside as a thin dusting of flour to keep the little cakes from sticking. The recipe says you may skip even that and use a wet cloth instead. The grain that would one day become this cake, almost entirely, enters its origin story as an afterthought that could be replaced with a rag.

The griddle cake, royal version — māṣa dosaka.

Pour the pure black-gram paste slowly onto a well-heated griddle; spread it in a circle, and roast it in ghee. This is the celebrated dosaka*, delightful —* rājabhogya*, fit for a king.*

Fit for a king. The gram cake was not a rustic thing waiting to be enhanced by rice. It was indeed the celebrated version that the king was proud to set on his own leaf.

The griddle cake, blended version — caṇaka–taṇḍula dosaka. And only here, in a recipe of its own, does rice enter the batter at all, paired not with black gram but with chickpea:

Mix chickpea flour with ground rice flour; season with asafoetida, cumin, and sindhuvāra*; grease the griddle evenly with oil, spread the batter across it, and cook the fine* dosaka well on both sides.

So in this king’s kitchen, the two grains, rice and urad (black gram), that now define the famous fermented pancake of the south never once met on the same stone. Rice was married to chickpea, in a cake of its own. The urad and rice union that we now know as the very definition of the dish had not yet come together in the twelfth century.

We at least know that the griddle cake had been seen long before the king wrote it down. It surfaces as a passing observation in the old Tamil verse of the Sangam age (the classical period of Tamil literature, roughly the third century BCE to the third century CE), perhaps as early as the first centuries of the common era. The verse notes and names it, as an observation in the course of someone’s day. But it never mentions the method by which it was made. What the king set down in his book was the first time anyone explained how the thing was made. But a recipe does not grant you ownership.

For a long time the received wisdom held that the people of the subcontinent must have borrowed the trick of fermenting and steaming from somewhere abroad. You can let that go as the king’s own recipe contains both moves: the batter is kept until it sours, the cake is cooked by steam. Nobody had to import fermentation to a wet monsoon kitchen. We know that a bowl of ground lentils left out overnight in that setting is bound to turn tangy by morning on its own. And while regions still squabble over whom the cake belongs to, the squabble dissolves the moment you notice the dish is older than the record of it, and older than any of the borders now fighting over it. The cake belongs to everyone. The recipe has an address.

What none of that explains is the thing that did change. Not the method. The ratio.

Here is where the cake stops being a curiosity and becomes a record.

The transformation was not the addition of rice but the inversion of the proportion. In the beginning, black gram was the master. Rice, when it came at all, was a servant brought in as a mere whisper, for a little crispness and to help bind, the way you might add an egg. The version eaten across the south this morning is the reverse. Rice with a whisper of gram, three or four parts rice to one, the gram demoted to a mere agent of fermentation in what is, by now, essentially a cake of steamed rice. Somewhere across eight centuries, the servant became the master.

To see how that happens to a food, you have to look at what happened to the ground it grew in, because rice is not a humble grain. It is thirsty; it wants flat flooded fields and a great deal of labor; and it was, for most of southern history, expensive and was treated as currency. It was the crop of a narrow stratum: the people on the best-watered land, nearest the rivers and the temples the rivers fed, who were mostly exempt from taxes by a special grant called brahmadeya. The exemption had a name, irai-ili (irai, tax; ili, devoid of). The medieval tax records lay the hierarchy bare. Rice grown in abundance dictated by the genius of the land and the water that always stood underneath left enough surplus to experiment. Everyone else, which is to say nearly everyone, lived on the dry grains that ask little and forgive much: the millets, and the pulses. The grain on a plate announced where its eater stood.

Then the land tilted, and the cake tilted with it. In the great river deltas, the Kāverī (the great river of the Tamil country, rising in the Western Ghats and reaching the Bay of Bengal) above all, rice farming concentrated and intensified, held in small parcels by households best suited for it. Where rice ruled the field, it began to rule the batter. And in doing so the cake turned into something you could read: a luxury and discretion, legible in flour.

Consider what the rice dominant white version is, in plain economic terms. It takes the most expensive, most water-hungry grain that is used as a tax currency in a household and commits three or four parts of it to a single cake. Then the cake is steamed into something airy to be eaten by the dozen at breakfast. For a household that grew rice as a fraction of its land precisely because rice was the costly specialization, making the white cake every morning would be lighting money on fire. My own people farmed for as far back as anyone can name. They grew rice on a sliver of their land, only what the water justified, the rest in dry grains. The rice on the table was only for festivals. They did not make the white cake at home, and not because they could not. Because it would have been foolish, and a farming family endures for centuries precisely by not being foolish with its scarcest grain. A family I knew of came from the delta, from exactly the kind of small, well-watered, rice-rich holding the old order rewarded, and their table carried more rice as a natural consequence of proximity to standing water. Both households, to this day, make the old gram-heavy cake. Neither will make it without a handful of rice, for crispness and hold. They keep rice exactly where the old world kept it as a functional minority, earning its small place by what it does for the texture, never permitted to take over. But the difference is, in one household it is a necessity and in the other one, it is just a practice to honor tradition.

Rice climbing from that small fraction up to three-quarters of the batter and going from binder to the body of the cake didn’t happen because cooks discovered it tasted better.

It happened because rice got cheap. So the increase isn’t a culinary development. It’s an economic one.

So the white cake was never merely refined. It was a conspicuous announcement by a household that it had so much rice it could afford to use it liberally. The whitest, most rice-forward version was made most freely in the temple kitchens and the prosperous households around them. This version earned a name and a pedigree that was worth aspiring to. Not because anyone schemed it so but as an accident of arithmetic. That is how class works most of the time, not by conspiracy, but by the patient accumulation of innocent choices into a structure nobody chose and everyone inherited.

By now you have very likely guessed what it is. The cake of the king’s encyclopedia, the iḍḍarikā that began as black gram and ended as rice, is the one the world now calls the idli, the soft white disc of the South Indian breakfast, and its crisp companion the dosa, sold today on counters from Chennai to London to New Jersey. I held the modern names back only so you would meet the cake as its makers did, and feel the distance it traveled. Now that you know it, watch how the journey ends, because the ending is the strangest part of all.

For centuries the white cake stayed mostly where rice was cheap enough to squander, in the temples and the households near them, gated not by any rule but by a price. Then the price broke, and it broke starting with colonial actions. The delta had been engineered for rice before, by the Cholas, for revenue. The British re-engineered it for the same reason. Cotton was commissioned to build on the Kāverī and Coleroon (the Kāverī’s northern branch) rivers from 1828. The Coleroon anicuts (diversion weirs thrown across a river to feed irrigation channels) were built in 1835 and 1836. Kallanai, built by Karikala Chola 1800 years ago, was remodeled and renamed the Grand Anicut. Cotton’s work would lead to increasing land revenue through agricultural improvement and the commodification of rice. Rice was the grain the revenue system could actually count as it was for a thousand years. A grain that already functioned as the unit of account is the grain a revenue state wants more of.

The British were doing arithmetic the Cholas had already done.

Then comes a very consequential twist of fate. Burma was annexed by the British empire in 1852. The Irrawaddy delta in Burma was cleared and transformed into the world’s largest rice exporter, financed substantially by Tamil moneylenders and cheap Indian labor. The Konbaung kings had held paddy in granaries and forbidden export. The British opened it.

Then in 1876, monsoon failed like never before in a long time. Famine hit very hard in the rain fed interiors of south India. The irony was rice was available in abundance but was left to follow the free market. The price went up by the multitudes and the grain went past the starving to the ones who could afford it. 5.5 million people perished needlessly.

This white cake witnessed all of that silently like ones who consumed it.

After independence, the canal and the ration card finished the work, pouring cheap rice into kitchens that had never had it in quantity. The gram-and-millet plate that had fed the subcontinent for millennia became, inside a single generation, the food of the poor and the past. And in the very same motion, the cost-barrier that had gated the white cake dissolved. When rice is cheap, whipping it into a breakfast cake is no longer a declaration of anything. The luxury stopped being a luxury.

The vehicle that carried it the last mile was, fittingly, homegrown: the South Indian “tiffin” restaurant (tiffin being the light meal taken outside the main midday one, and by extension the shops that served it), temple-trained Brahmin cooks and merchant-caste entrepreneurs riding the new urban migration, took the pale cake out of the private kitchen, standardized it, and set it on a counter in every Indian city, and then in every city the Indian diaspora reached. The empire made it affordable; local enterprise made it available. The accident opened the door; the restaurateurs walked the dish through.

And it should be said that the cake did not conquer on status alone. It spread, too, because it is good: light, digestible, fermented into something nourishing, forgiving to cook at scale, equally at home as breakfast or an invalid’s meal or a festival offering. Status got it noticed. Quality got it kept. The cake earned its universal life as much as it inherited it.

The colonial accident impoverished the diet and democratized the delicacy in the same motion. That is not a contradiction to resolve. It is the irony to sit inside.

So when someone hands you the palest, fluffiest, most rice-pure version and calls it the authentic one, they have it exactly backwards. The dark, dense ancestor which was the king’s black gram cake is the original and is still made in both kitchens. Its ancient status dissipated, it sits demoted quietly behind the new one. What went out to the world was the aristocrat’s recipe, finally cheap enough for everyone. In that journey this cake had witnessed the remarkable history of the Kings, the people and the land: colonizer and colonized. I think about that whenever the steam comes off the plate. The cake is not a humble food that got fancy. It is a common food that got conquered, switched its grain on the way up, and in time, landed back in the common table.

From the King’s side,
It travelled through all
Many an eon
Changing colors
But soul intact
Gated by a few
By accident alone
It finally broke through
Witnessing untold sorrow
During that sojourn
Back to everyone
Gently demoting
Its darker cousin

Subscribe now

Sources.

The three recipes are from the Mānasollāsa of the Chalukya king Someśvara III, c. 1130, ed. G. K. Shrigondekar, Gaekwad’s Oriental Series 84 (Baroda, 1939), Vimśati 3, ch. 13, vv. 1553–1555, 1561–1562 and 1562–1563; other recensions number and order these differently.

The name is older still: iḍḍalige*, in the Kannada* Vaḍḍārādhane*, c. 920.*

On the brahmadeya grant, tax-exempt tenure and the wet and dry land categories: Noboru Karashima, South India History and Society: Studies from Inscriptions A.D. 850–1800 (Oxford, 1984) and South Indian Society in Transition (Oxford, 2009).

On the re-engineering of the Kāverī delta and the colonial reordering of grain in the Tamil countryside: David Ludden, Peasant History in South India (Princeton, 1985) and Christopher Baker, An Indian Rural Economy 1880–1955: The Tamilnad Countryside (Clarendon, 1984).

On the opening of the Irrawaddy delta as a rice frontier: Michael Adas, The Burma Delta: Economic Development and Social Change on an Asian Rice Frontier, 1852–1941 (University of Wisconsin Press, 1974).

On the famine of 1876 to 1878: William Digby, The Famine Campaign in Southern India, 1876–1878 (1878), and the Report of the Indian Famine Commission, 1880.

With thanks to Harshita Saxena, whose Substack Epicurean Dispatch and the references she sourced for it struck a chord with my own interest in precolonial Indian cuisine, and inspired the sourcing for this piece.

Epicurean Dispatch - My Spicy journal

All about the journey of Indian spices. From vintage cooking methods to modern food techniques.

By Harshita saxena

My friend had come back from his latest trip overseas. It was essentially a cultural immersion tour, as he had some roots to that country. I had invited him over to one of my dinners for friends, where I test my novel ideas on them. He is from Louisiana originally, and I had spent countless hours preparing this gumbo especially for him, along with the rest of the table.

I greeted him with the usual, “How was your trip this time?”

He started talking about all the Michelin-starred places he had been and how amazing they were, and about his encounters with the chefs. He felt very privileged to have eaten there and to have met some of them. I followed up by asking if they had been to any hole-in-the-wall places, anything recommended by locals. That is what I usually do. But he outright dismissed that and said they had selected as many Michelin restaurants as they could get reservations for, and the rest were the most popular restaurants in the city. They had checked most of them off their list. I was a little taken aback and asked him if having so many meals at these restaurants would be overwhelming, to which he said that is what they do on every trip. He continued to talk about the plating, the caviar on desserts, the liquid nitrogen fog that had poured over his table between spoonfuls.

Having several tasting menus in sequence over several days is not for the faint of heart. It takes a particular mindset. The kind where the inner psyche is using food for a different purpose altogether.

I have to admit I had cooked the dish to get a certificate, to validate myself in front of a native who had eaten the real thing many times over. He was collecting stars instead of flavors. I was collecting the approval of a native authority. All of us have checked a restaurant off a list at some point, or stood in line for hours at an Instafamous bakery.

Some of us do it quite a bit more than we would admit to ourselves, my friend among them. The reason he is so obsessed with the prestige, the status, the theater of it all, is that for him food is a currency. A currency of social capital, a way to show the world his standing in it and a sign of what he has accomplished in life. The food itself comes second.

Where you eat, who serves you, who the food was meant for, how scarce it is, all of that earns more value than what is actually served. This is the layer the elite establishment has built into an edifice. Once you have arrived at a station, taking part becomes an inevitable exercise in inclusion. Some do it ceremonially. Others embrace it fully. Scarcity, and the use of food as social capital, has been with us since as far back as the invention of the ledger[1]. Institutions and social media, built on that old scaffold, may have taken us into unforeseen territory.

You usually do not hear about that one dish that blew their mind, or the one sauce that froze them in time. Instead you hear how hard it was to get the reservation, or how much it cost. Some of them design the whole trip around those reservations.

I had started the day gathering the ingredients: the trinity of pepper, celery, and onion,[2] the sausage, the shrimp. I began with the roux, stirring until it turned mahogany, which is where the time and the care go. The shrimp, the sausage, whatever meat is on hand, all of it is about making something from what little there is, the way people stretched thin have always cooked. The dish is elevated by what goes into the beef bouillon, the Cajun seasoning and the aromatics, and by the star of it, okra toasted in bacon drippings. It takes the better part of a day, which is why in Louisiana it is a Sunday ritual, cooked slow before the game. I had spent half a day at this preparation, but it did not feel like it. I enjoyed every hour, because I was reliving the lives of folks down south, making an ordinary soup extraordinary.

Yet this half day of pleasure and inheritance found no column on the checklist. It cannot score a soup. There is no star to earn, no reservation to win, no scarcity to brag about, so the whole thing gets filed under comfort, the lesser shelf, the thing you supposedly settle for when the star table falls through.

But as my friend continued to taste the gumbo, the flavors started to hit him, waking him from the Michelin stupor with memories of Sunday football in Louisiana. He looked at me, startled, and said, “You cooked gumbo for me, and it is actually pretty good.” The flavor was just the vehicle. It took him home. The thing that takes you back, the thing you return to, is the one thing the whole system has no way to value, because going home earns no star. The ledger does not register the equivalent of a grandmother’s hug, the thing that is supposed to outrank anything.

None of us is a monolith, and neither is my friend. He is predominantly a checklist eater, but in him is also a person who wants to eat for his soul. The soup that woke him up proves it. We are all the same as humans. We carry these as layers or stacks within us, and it is up to us how we stack them.

[1] Not a figure of speech. The earliest known writing, proto-cuneiform pressed into clay at Uruk around 3300 BCE, is almost entirely accounting, and most of what it counts is food: measures of barley, loaves of bread, jars of beer issued as rations. Administrative records make up the overwhelming majority of the surviving tablets. The ledger was invented to keep track of who was fed and how much, which is to say the counting of food is older than the writing of anything else.

[2] Onion, celery, and green bell pepper, Louisiana’s version of the French mirepoix, with the pepper standing in for the carrot that did not grow well there. The vegetables are old and the name is not. Paul Prudhomme popularized “the holy trinity” in the late 1970s, and it was standard usage by the early 1980s.

Subscribe now

I was not eating an exceptionally made dessert but actually reliving the emotion.

I used to be puzzled when my kids and their cousins would adore the dishes my mother in law made. I would say ‘I have eaten better’. They would ignore me with an emphatic vote for their grandma as the best chef there ever was. Granted, she was good at what she cooked, now I realized that what they were feeling was this overwhelming selfless love of hers in her dish. The dishes were literally their grandma’s warm loving hug. Their voting her the best chef alive was never a verdict on technique. Her dish was bundled with the oxytocin of being loved [3], and my mastery cannot out-cook a hormone.

It does not have to be a grandma always. My sons love Butter Chicken/Chicken tikka masala. I have cooked that dish for them a zillion times. I think I have mastered that dish and have been told so by many. But my son has a different reference point. Before I started cooking or had time to, he in his early years had eaten that dish in a restaurant in town. I have known this chef well for the last two decades and he is good. For my son that was a food which gave him consistent pleasure and comfort. The emotional warmth it created will not let my culinary excellence surpass it. I had a reference point of the dish of how it should be based on how it is made back in India. But that did not matter to my son. This is when I realized that passport is not the palate.

I had the same experience in the famous Michelin-starred restaurant, Semma. When my family and I were eating there, we felt a little deflated as most dishes were distant cousins of the versions back in India. But when we looked around we were flabbergasted to see Americans of all ethnic origins chomping at it with almost indecent pleasure. The chef had evoked an emotion that bonded them to that food. They did not have a reference register to measure against and they would not care. Who am I to judge them. Passport is not the palate. And this is the Ship of Theseus again, the one I wrote about with congee. A dish crosses an ocean, its planks swapped one immigrant kitchen at a time, and it stays the same dish for whoever the emotional slot still fits. For me the swapped planks read as loss. For them there was no distortion to notice, because they had nothing to measure it against.

Personally I have a few restaurants whose dishes evoke that response in me. One of those restaurants is one called Bacanora in downtown Phoenix. The mango prawns in that tangy sweet and spicy pineapple salsa is an emotion for me that brings the same joy of my great grandma’s hug. If no one was looking, I would literally lick that sauce clean shamelessly [2]. Their thinly sliced scallop crudo feels like tangy tender coconut taking you to an island paradise with every bite. I heard from the waitress that the chef had won a James Beard award. But my palate was on a blind date and fell in love with him.

When a chef can evoke an emotion like that, it is extraordinary. It’s like a stranger who feels like a loved one.

That is an amazing feat. One other restaurant where I felt like that was Quintonil in Mexico City long before Michelin had to be bribed by the city to come and check them out. My palate ranked them before Michelin ever did.

Just last week, my family and I were at our favorite restaurant, Rasika in Washington DC. We were visiting that restaurant after a few years. We still have fond memories of some of the exquisitely executed dishes. But when we had the dishes, we all felt let down. We felt that care and attention to those dishes were gone. It felt like a lover who had gone cold. We then learned that the group that has always owned them was actively scaling up, opening restaurant after restaurant. We have heard this story before. Expansion for economics and the food tends to flatline. Maybe it is just us and they are focused on recruiting more naive palates. They may win new customers faster than they lose the ones who would have come home again, and for a while the math looks fine. But soon that may flatline as the craving to come back again dries up without that emotional hook [1]. There are only two ways to grow. Centralize the cooking and every location becomes reliably the same, which means reliably fine. Or leave each kitchen to itself and the good nights shine but there will be bad ones. One path takes away the height, the other takes away the certainty. A craving needs both. This makes me tremble to think what would happen to Bacanora if they expand. I cannot last a month without that emotion. I know we all have one of those places.

We all need that anchor in food!

Subscribe now

A note on the three chemicals, for the curious.

[1] Dopamine is the wanting. It is the anticipation, the pull toward the door, the craving that makes you return before you have tasted a thing. It builds when a reward reliably shows up, and it fades when the reward stops landing. This is the chemistry of coming back.

[2] Endorphins, the body’s own opioids, are the liking. They are the actual pleasure of the bite, strongest with fat and sugar and salt. Wanting and liking are not the same system, which is why you can crave something you no longer enjoy, and enjoy something you never crave.

[3] Oxytocin is the bond. It is the molecule of nursing and touch and being cared for, present at the very first meals of a life. It is why a dish can carry a grandmother’s love and why no amount of technique can out-cook it.

Then came my one savior, congee with fried onions, an egg and shrimp paste. A starving body does not taste, it audits. The dissolved starch was fast energy, the egg and the shrimp paste were the reserve. That bowl fed the fueler in me, not the palate. It sustained me until my flight the next day where my in-flight meal was waiting to rescue me. All this that happened thirty-three years ago flashed in my mind when I was offered congee for breakfast on my recent Cathay Pacific flight to Hong Kong on my way to India.

Congee is a dish where the rice is cooked with a disproportionate amount of water or stock long enough to make the grains swell and rupture, dissolving into the cooking liquid itself. This becomes a new fluid where the grain and liquid merge into one. This is unique for congee as other methods tend to preserve the grain. That process is accelerated by mechanical stirring along with patience until the long moist heat fully gelatinizes the starch. But the dish comes deliberately bland to the table as a plain canvas where you paint with condiments and garnishes.

This dish is ancient and has its own long journey. But, before we move forward, I would like to take you back to ancient Athens. The Athenians are known to have preserved for centuries a ship that Theseus sailed back from Crete. They did that by replacing the planks as they rotted with new timber. This posed a question for the ancient philosophers. What happens if every plank is replaced? Was it still the Ship of Theseus?

The philosophers never settled the question. In fact, Plutarch asked this question centuries later but did not answer. But we are going to answer that question using Aristotle’s form/matter distinction, even though he himself never posed this question. Based on his principle, even if the vessel has its planks fully replaced, it persists if it preserves its form and purpose. So, the ship changed every plank and kept its name. Congee ran the same experiment in reverse. For three thousand years the bowl kept its planks. The drowning of the grain in water, the patience, the swelling and the rupture, none of it changed. What changed was the name, and it changed at every border. Zhou in China. Juk in Canton. Okayu in Japan. Kanji in the Tamil country. Bubur in the islands. And congee in English, which is the Tamil word coming back in borrowed clothes. The ship kept its name and lost its body. This dish kept its body and let the name wander.

Congee is zhou in Chinese, which is its original name, and the earliest recorded reference to it is around 1000 BCE during the Zhou dynasty. The Book of Rites (Liji) describes zhou as part of everyday life, recommended for the elderly and the sick. It was also the rescue dish for the starving during the famine years. Rice gruel was a universal dish spanning many cultures, and India had its own version as well. The ancient texts in India mention the gruels in many forms based on the rice to water ratio: yavāgu at 1:6 rice-to-water, vilepī at 1:4, peyā at 1:14, maṇḍa the strained liquid of a 1:14 boil.

In the fifth century BCE, the Buddha named the advantages of these rice gruels, and with Buddhism’s spread into China, the monastic breakfast entered the Chan monasteries as a daily code with a fixed hour for it; the Tang temples became standing infrastructure, serving zhou to the poor and the pilgrims as routine charity, with no famine required; the Laba zhou became an Indian story celebrated with a Chinese porridge on the eighth day of the twelfth month. That is how zhou’s purpose was sanctified by scripture.

So, zhou spans the entire status ladder in society. It was enjoyed throughout the rest of Chinese history as a dish of the emperor, the imperial banquet congee, besides being the humble porridge for the peasants and the sickly. Crowned with abalone, dried scallops or century egg with pork, it could grace an emperor’s table, while the same bowl with a scatter of scallions and white pepper fed the peasant. My Cathay bowl was the imperial version, congee in pork broth with prawns, caramelized scallion and half an egg. The water was promoted to broth and the peasant’s raw scallion was promoted by caramelizing, but the ratio itself was never touched. The dish climbed the ladder without breaking its own rule. Korea makes it explicit that juk is food for recuperation, a delicacy, or famine food depending on preparation, all three registers in one dish. In the meanwhile, India exported the sanctity via theology that helped zhou climb the ladder in China and beyond but did not keep any for its own bowl. It supplied the blessing but did not apply it. The Tamil word kanji is still a dish for the poor and the sick.

Coming back to the name congee, this is where it gets personal. The dish is never renamed anywhere but just code-switched based on whether it appears in the English menu column like it did on my in-flight menu. The word congee is English’s own inherited name from the Tamil fossil kanji, a trader’s mispronunciation carried by the colonial trade through India and out to other Asian ports. This came home to me after I landed in Chennai where I had Chinese congee on the menu at my hotel. But I needed to custom order the indigenous kanji if I wanted one. Colonial coding runs deep and kanji is still struggling for a higher rung. But the kanji that arrived brought me the warmth of my nanny from my childhood in India. She ate it every single day and called it her superfood, the one thing that could keep her going all day. The world’s ledger placed her bowl on the bottom rung. She carried it as her strength. That bowl was another emotion altogether.

But one thing is for certain, on that crazy day thirty-three years ago at the Hong Kong airport, this dish stayed true to its form and purpose even if it takes different names by rescuing this starving human in dire circumstances.

Also, let another thought sink in. Do you think any Michelin inspector has taken this view on congee before deciding to judge one or not to judge at all?

It marches silently
Through eons weathered
By all we have done
It speaks to us
More than we ever can
While we fight to own it
It quietly owns all of us

Subscribe now

Let that sit for a moment. All that Instagram hype, all that viral energy, all those carefully filtered photographs of dramatic plating and towering desserts — and seventy percent of the people who arrive driven by that content, eat once, and never come back. Not because the city ran out of restaurants. Because the experience did not deliver what the perception promised.

The data is even more specific than that. Sixty-nine percent of guests are one-time visitors. Only eight percent achieve what the industry calls regular status. The annual churn rate costs each location approximately $375,000 in lost revenue every year. And yet sixty-five to eighty percent of restaurant revenue comes from repeat customers: the regulars, the returnees, the people who came back because something true happened to them at a table.

The Instagram restaurant has built its entire proposition on the visitor who was never going to return. It has optimized every decision — every dish, every dramatic plating, every loud music choice, every cramped table arrangement designed to maximize covers — for the seventy percent. And in doing so it has systematically destroyed its relationship with the thirty percent whose loyalty would have sustained it.

The Instagram restaurant gets the first visit. The mint sorbet gets photographed and posted before anyone has tasted it and found it wanting. The followers see the post. They book a table. They get the mint sorbet. They reach for water. They post anyway because they came to post not to taste. They don’t return. New followers see the post. The cycle continues.

It is a perfect machine for generating covers. It is a perfect machine for destroying a restaurant from within.

Avartana, with no algorithm behind it, is full of the people who came back. Repeat guests spend sixty-seven percent more per order than first-time visitors. The rasam that made a table fall silent. The bitter gourd that stopped a sentence mid-word. The staff member who remembered your name. These are not Instagram moments. They are return visit reasons. They are the rack and pinion of a restaurant that will still be here in ten years when the Instagram restaurant has cycled through its algorithm and closed.

The placebo palate arrives once, photographs, posts, and leaves. The palate that has been fed the truth just once, completely, comes back.

That is not sentiment. That is the business model that works.

This is the placebo palate in its most complete form.

Jean Baudrillard argued that contemporary society has replaced reality with simulacra: copies so far removed from their original that they no longer reference anything real. They reference only each other in an endless loop of signification detached from truth. He was writing about media and consumer culture in the late twentieth century. He could have been writing about any Instagram-famous restaurant in Chennai in 2026.

The posts reference the dishes. The dishes reference the posts. The diner arrives to photograph the experience of photographing the experience. The mint sorbet exists not to cleanse the palate but to fill the gap between courses in the content sequence. The most provocative dish is on the menu not because the kitchen has mastered it but because it generates the kind of caption that generates the kind of engagement that generates the next reservation.

But let us be honest: food has always been performed. This is not new. The royal feast, the temple prasad, the aristocratic table — all used performance to amplify what was genuinely worthy. The Mughal court’s endorsement of biryani didn’t corrupt biryani, but it gave it resources, refinement, and reach. The temple’s sanctification of certain dishes embedded them into cultural memory deeply enough to survive centuries. The performance served the dish. The dish had to be excellent to earn and hold that elevation. Reality and perception moved together.

The corrective mechanism was built in. The same palates returning day after day. Memory. Standards. Consequences. A bad dish at a royal table faced an immediate and sustained reckoning.

Social media destroyed the corrective mechanism entirely. The Instagram audience rotates perpetually. New diners arrive driven by old posts. The collapsed taco gets photographed before it collapses. The mint assault gets described as bold and adventurous. The five stars accumulate. The kitchen receives no honest signal. The dish never gets fixed. The simulacrum perpetuates itself.

What was once ninety percent reality and ten percent performance, the performance in service of the reality, is now ninety percent performance and ten percent reality. The reality serves the performance if it serves anything at all.

And the scale is catastrophic. A royal endorsement reached hundreds over years. A viral Instagram post reaches millions overnight. The perception crystallizes into received truth before anyone has eaten the dish enough times to know whether it deserves it. The placebo palate gets institutionalized before the corrective experience can accumulate.

But the placebo palate did not arrive alone.

In India it arrived with a partner: the discounted palate. Zomato and Swiggy built their dominance on investor-subsidized price wars, training an entire generation of urban Indian diners to evaluate food by discount percentage rather than quality. Fifty percent off. Free delivery. Buy one get one. The algorithm rewards what travels well in a plastic box in forty minutes. Everything else becomes invisible.

What travels well is maida. Sugar. Sodium. Ultra-processed, shelf-stable, packaging-optimized food engineered for the delivery radius not the palate. What does not travel well is aappam, its lace edges collapsing before it leaves the kitchen. Ottu shavige — pressed fresh, eaten immediately, dead within minutes of leaving the pan. Vatha kuzhambu made the real way — the slow simmer, the manathakkali, the depth that lives in time — indistinguishable in delivery from the shortcut version that took twenty minutes instead of three hours. The rice appalam garnish, that magic crunch, arriving soggy and defeated.

The delivery platform didn’t set out to destroy traditional food culture. It set out to make food convenient. But convenience, scaled to the size of an algorithm and subsidized to the price of a discount, becomes its own kind of violence. Quiet, impersonal, nobody’s fault, everybody’s loss.

What cannot travel gets tethered to the place it was made. Pani puri’s crisp shell absorbs moisture within minutes — dead before any delivery radius is satisfied. Ottu shavige must be pressed and eaten immediately. Real wasabi, grated from the rhizome and used within the hour, has been replaced almost everywhere on earth by colored horseradish paste because the cold chain that could carry the real thing makes it financially prohibitive outside Japan. These dishes cleared every quality gate. The logistics gate was the one that held them back. Refrigeration liberated food from geography for everything it could carry — and quietly left behind everything it could not.

The economic consequence compounds the cultural one. Restaurants participating in discount programs compress their margins to the point where quality ingredients become unaffordable. The race to the bottom selects for the cheapest possible version of a dish that can survive delivery. The craftsman who knows how to press ottu shavige has no economic model on the platform. The kari dosai chef at the Marriott breakfast station, carrying a tradition in his hands all morning, will never appear in an app.

And then the health cost arrives. Not immediately. Downstream. Invisibly.

India has over a hundred million people living with Type 2 diabetes, the second largest diabetic population on earth. Urban India’s rate is nearly double rural India’s. The variable is the city diet. The shift happened within one generation — the same generation raised on Zomato discounts and Instagram food culture. Ragi mudde and ragi simili urundai are among the lowest glycemic index foods in existence, metabolic health protocols disguised as humble food. They are now considered poor people’s food and disappearing from urban menus. The Ayurvedic framework of six tastes (bitter, astringent, sour, pungent, sweet, salty), each with specific metabolic function, calibrated over centuries of experiential wisdom, replaced by delivery menus optimized for sweet, salty, and umami because those photograph well and get reordered.

The dahi rice that ended every meal in that Coonoor villa has probiotic, acid-neutralizing, serotonin-inducing properties and anticipated modern gut health science by a thousand years. The adai consumed as emotion at Karpagambal mess, complete protein, iron delivery system, metabolic calibration disguised as an extraordinary college meal, is leaving commercial menus because it cannot be flipped in seconds on a thawa.

If you choose food only by what travels well and fast to you, you may not travel far into your lifespan.

And when the food goes, everything encoded in the food goes with it.

The recipe is the most durable form of oral history. The meal is where language, ritual, relationship, and identity converge daily. When a grandmother stops making aappam because her grandchildren grew up on Swiggy and never learned to use the appachatti — a recipe dies, a language loses a word, a body loses its medicine, and a civilization loses a memory it will never recover.

The Toda people of the Nilgiris did not choose this. Their pastoral culture, their sacred buffalo, their dairy tradition that was also a cosmological system — taken from them through colonial land seizure, forest reserve classifications, the systematic displacement of everything their food culture depended on. They had no agency in their erasure.

Urban India is making the erasure voluntary. One discount order at a time. One Instagram reservation at a time. One untested sorbet on a tasting menu that nobody honest has reviewed at a time.

The mechanism is different. The loss is the same.

And this is not only an Indian story.

In every city where a delivery platform arrived with investor subsidies and discount promises, the same sequence unfolded. The traditional breakfast culture receded. The grandmother’s recipe went unlearned. The indigenous ingredient disappeared from menus. The cuisine that could not survive forty minutes in a plastic box was quietly discontinued. The algorithm rewarded what photographed well and delivered fast. The rest became invisible.

The full English breakfast replaced by meal deal sandwiches. The French café au lait and tartine displaced by drive-through coffee. The Japanese ichiju sansai morning meal giving way to convenience store onigiri. The Sichuan pepper dialed down in Chinese restaurants from San Francisco to Sydney to satisfy the dominant demographic’s preference for sweetness over complexity. The Som Tam losing its spine in Bangalore, in London, in New York, and wherever the market decided that the tension was too much and the sweetness was enough.

The placebo palate has no passport. The discounted palate has no geography. Instagram has no borders. The simulacrum of dining experience is as present in Paris as in Chennai, in Lagos as in London, in Mexico City as in Mumbai.

Every food culture on earth has its Toda story: the indigenous, the marginal, the unchampioned, the erased. Every city has its Laghu Ahara, the traditional food culture that walked past invisibility into extinction while the algorithm celebrated something louder and more photogenic. Every family has its three dishes, held in the hands of women who learned from women who learned from women, retreating slowly into private homes, their names surviving on menus long after the dishes themselves have forgotten what they were supposed to be.

Bangalore is not unique in this. Chennai is not unique in this. The world is simply not yet honest about it.

So what do we do.

We do not fight the platform. We do not refuse the market. We do not legislate the algorithm or ban the discount or close Instagram. None of that is the answer and none of it would work.

We do what Indian Railways did with the toy train.

We find what is irreplaceable. We tell its story so precisely, so honestly, so beautifully that the diner feels the loss of not having known it sooner. We name the dish correctly: not just dosa but khali dose, not just pudding but elaneer brûlée with spiced jaggery, not just minced meat but kari dosai made by hands from the birthplace of the dish, not just a gravy but vatha kuzhambu slow-simmered with manathakkali that nobody forages anymore and somebody must, with the rice appalam garnish that makes it worth becoming a food thief for. We celebrate the craftsman publicly: the chef at the breakfast station who has been carrying the tradition in his hands waiting for someone to ask. We build the mythology the way MTR built the mythology of rave idli: with a story, a specific geography, an identity so rooted that no delivery platform can replicate it and no discount can substitute for it.

We reclaim the signal. One honest review at a time, one precisely named dish at a time, one food writer who has been telling the truth since 2010 at a time. We make the genuine article feel inevitable rather than unreasonable. We make the diner feel complicit in something beautiful when they seek it out and complicit in something lost when they don’t.

We revalue the transaction. The true cost of the discounted meal includes the downstream health cost that arrives in hospitals and pharmacies a generation later. Full price for ragi mudde is cheaper than metformin. Full price for aappam made with overnight fermented batter, with the floating jiggling yellow of the egg in the hollow, the crisp edges, the soft center soaking the stew, is cheaper than the gastroenterologist. Full price for vatha kuzhambu made with real manathakkali and three hours of slow simmer, finished with the magic of a rice appalam, is cheaper than everything the shortcut version costs you that you will not know until much later.

We educate the palate. Not through instruction or superiority but through experience. One meal allowed to be completely itself. One dish that arrives and does what only that dish can do — consumes you as emotion before you have even finished eating it. The Chettinad curry arriving in waves. The kari dosai warping past and present into a single bite. The crack of a spoon through a spiced jaggery crust on a Tuesday afternoon in Bangalore. The aappam on a Sunday morning that was always a prelude to something restful and right.

We remember that the appetite for the real thing is always larger than the market assumes. The room at the Marriott breakfast buffet, full of people who had walked past that keema without a second glance, became a room that did not want to stop eating once someone asked the right question. The dish was always there. The craftsman was always there. The hunger was always there.

The placebo palate was educated into existence one Instagram post at a time. It can be uneducated the same way — one honest meal, one true story, one dish allowed to be completely itself at a time.

Every cuisine worth saving has its rack and pinion: the thing that cannot be replicated, cannot be delivered, cannot be discounted without destroying. The appachatti that makes one aappam at a time. The manathakkali that requires someone to forage and dry it with patience. The fermented batter that cannot be rushed without breaking. The kari dosai that requires the hands of someone from the birthplace of the dish. The rice appalam that arrives as the final argument of a vatha kuzhambu that took three hours and deserved every minute. The elaneer brûlée with its spiced jaggery crust that knew where it was and what it was made of and had nothing to prove.

Find that. Protect that. Tell the story of that until the diner understands why the mountain cannot be climbed any other way.

The Toda dairy is waiting. The Laghu Ahara soul is waiting. The aappam, the adai consumed as emotion at Karpagambal mess, the vatha kuzhambu worth becoming a food thief for, are all waiting. Not gone. Not lost. Only underfed.

And in a quiet restaurant in Chennai that the algorithm has largely ignored, a chef is preparing rasam tableside for a room of people who came back. Not because they photographed it. Because it was true. Because truth — fed once, completely, without apology — is the one thing the placebo palate cannot simulate and cannot replace.

The palate was never lost. It was only underfed. Feed it the truth once — just once, completely — and it remembers everything.

Subscribe now

I want to be fair, because the flatness is instructive precisely because it’s a good version of the thing. Mon Ami Gabi is a chain, run by a large group, engineered to do exactly what I felt: hit the mark every time, never miss, never reach. The bentwood chairs are there, the chalkboard, the rolling wine trolley, the votive, the leather menu, the tricolour near the door. Every signifier present and correct. “Minimum standard, reliably met” is a real achievement at that scale. It is hard to make a thousand competent steaks a week. But it is an operational achievement, not a culinary one, and you can feel the difference, because the room is pitched at the register of a place where someone cooks for you while the kitchen is operating at the register of a place that executes a spec. The candlelight was writing a check the kitchen was never staffed to cash.

I have a word for what was being sold to me instead of dinner, and the word is performance. Better than ceremony, because it carries two meanings and the gap between them is the whole story. Performance is the show: the trolley rolled out for you to see, the theatre of Frenchness aimed at your judgment to manufacture a verdict the food didn’t earn. But performance is also how well a thing does its job, the plate delivering against spec. An honest restaurant keeps the two bonded; at a great sushi counter the show is the substance. Mon Ami Gabi is where they came unbonded. The theatre performed. The kitchen didn’t. And I was the audience.

So I started checking the claim against the one dish that can’t lie.

Order the steak tartare. It is the cleanest instrument on any French menu, because it strips every variable away. Raw beef, chopped by hand, bound with mustard and egg and capers and a few drops of Worcestershire. No cooking. No day-long reduction. No luxury ingredient hiding the arithmetic. The cost floor is a good cut and ten minutes with a knife. Whatever you are charged above that is not food. It is frame. Which makes tartare a kind of thermometer: stick it into any restaurant and read off, in dollars, exactly how much performance that room is running. So I followed one plate of raw beef across cities and tiers, and watched the number come loose from the food.

Start where I sat. At Mon Ami Gabi the tartare is $23.95. Walk it up. At Thomas Keller’s Bouchon (a Michelin-pedigreed bistro, the brand-name good version, a chef who truly reveres this food) it is $25. Hold those two side by side: the celebrated chef’s tartare and the chain’s tartare, the same raw beef, priced within a dollar. Whatever you pay for at Bouchon, the beef isn’t it.

Go up a floor, to the costume tier, where the room is grandest and the frame is sold hardest. At Balthazar (Keith McNally’s SoHo brasserie, the most famous French room in New York, all mirrors and red banquettes) the tartare is $26 as a starter, $35 as a main. The grander the theatre, the dearer the identical chopped beef.

Go up again, to the temple, and the dish does something revealing: it disappears. At Guy Savoy you cannot order a steak tartare at all; it survives only as a single seared bite folded into a tasting menu that runs about €415 in Paris and $455 in Las Vegas. And at the very top the absurdity completes itself. At Joël Robuchon in Las Vegas (three Michelin stars, the Chef of the Century) one of the house signature dishes is Le Rumsteack en Tartare et ses Frites à L’ancienne: steak tartare with old-fashioned fries. Raw beef and fried potatoes, the literal peasant floor of this entire essay, enshrined as a signature and served inside a dégustation that climbs toward $525 a head.

Then walk all the way back down, in the same city. Fifteen minutes off the Strip, at Ohlala French Bistro, a tiny chef-owned bistro out by Summerlin, the steak tartare is about fourteen dollars. Same town, same suppliers, same labour pool as Robuchon’s $525 rite. The owner, Richard Terzaghi, didn’t back into that price; he built the place on it, because when he arrived he found no honest French bistro in a city where everything was, in his words, “super expensive on the Strip.” He looked at the prices being charged for peasant food, understood they were a costume, and declined to wear it.

Here is the same plate of raw beef, top to bottom:

Read it top to bottom and the swindle reads itself. The same fistful of chopped beef costs roughly forty times more at the summit than at the floor, and it never improves once on the way up. Nothing is ever added to the cow. Everything is added to the performance.

And tartare is no fluke you can wave away as one odd line item, because the other peasant dishes tell the identical story. Escargots are snails and garlic butter, about as cheap as a protein gets: $12 at an off-Strip bistro, $26 at Balthazar. Twenty-six dollars for snails, because the butter is speaking French. Steak frites is the original peasant equation, a cheap cut and a fried potato: around $44 at Mon Ami Gabi, $56 at Bouchon. Run the thermometer through any of them and it reads the same: the price tracks the frame, not the food. Then back to the tartare, because tartare reads cleanest.

Two rooms in that table are doing something the rest can’t, and they are the reason the whole argument holds.

The first is Brasserie Zédel, a few steps off Piccadilly Circus. Down the red-carpeted stairs is a Belle Époque palace: marble, gilt, chandeliers, a vaulted Art Deco hall the size of a ballroom, live jazz most nights. It is more performed than Mon Ami Gabi, grander than Bouchon, more theatrical than anything in this essay short of the temples. And the tartare is £16; three courses are £19.95. Its founder, Jeremy King, said plainly that London needed a large, beautiful room serving great food at low prices, from prince to pauper, and then built exactly that. Zédel is the control that breaks the industry’s one excuse: the grandeur costs money, that’s why the plate is dear. Zédel is grander than almost all of them and cheaper than every one of them, which proves the performance was never the thing that cost the money. Staging the theatre is cheap, if you decide to eat the margin. What’s expensive is the decision to charge for it: the positioning, not the production.

The second is the whole country of France, where Zédel’s £16 palace is not an anomaly but the baseline. In Paris, at a famous bistro like Bistrot Paul Bert, the tartare is just a main course, and a whole lunch starts at €19. In Lyon a three-course bouchon dinner is €41. The bouillons (the grand, cheap Paris canteens like Chartier, the very model Zédel was built on) have served the working city in beautiful rooms at low prices for over a century. In France the honest plate in the handsome room is simply dinner. The costume isn’t for sale there, because there is nobody to sell it to: everyone already knows it’s peasant food.

Which is the part that turns a quibble into a scandal. Almost the entire French canon is peasant food. Not most of it. Almost all of it. Run the dishes and watch the aristocracy evaporate. Cassoulet is duck fat, beans, and cheap preserved meat, field food from Languedoc. Coq au vin is an old rooster too tough to roast, so you drown it in wine for hours until it surrenders. Boeuf bourguignon, daube, blanquette: cheap braising cuts and time. Confit isn’t a luxury, it’s a preservation method, how you kept meat alive before refrigeration. Pot-au-feu is boiled scraps. Ratatouille is what you do with a summer glut of vegetables. Bouillabaisse was Marseille fishermen cooking the ugly fish nobody would buy. Soupe à l’oignon is onions and stale bread, the cheapest things in the house. Steak frites itself is the cheap cut and the fried potato. Even the bistro as an institution is working-class: the word comes from cheap neighbourhood eating-houses, not from anywhere grand. The food the trolley and the votive perform reverence toward is, top to bottom, the food of people who had no trolley and no votive.

To be exact: France does have an aristocratic cuisine, and it is not an illusion. The haute cuisine of Carême and Escoffier was born in court kitchens and grand hotels, a machinery of consommés, mother sauces, soufflés, and pressed duck that demanded brigades of labour no farmhouse ever had, and when a temple charges temple prices for that work, the bill is at least honest about its ancestry. That tier is real, and it is not the subject of this essay. Look at what the trolley and the votive are selling in every room below the temple. It is not consommé. It is beans.

So the coding is inverted from the truth. French got coded in America as inherently fancy, special-occasion, priced-up, and the food being coded that way is duck fat and beans and tough old rooster and the fish nobody wanted. The performance isn’t dressing up truly aristocratic food. It’s dressing up peasant food and charging you for the costume. The cassoulet doesn’t know it’s supposed to be expensive; some grandmother in Languedoc made it to feed field hands cheaply, and three centuries later it sits behind candlelight at a markup because the word “French” got coded rich somewhere on the crossing.

And it’s worth being exact about how that happened, because a cuisine crosses an ocean through one of three doors, and the door it comes through sets its price forever. The first is the vernacular door: poor immigrants cooking their own home food for their own community, cheaply, because they are feeding each other rather than performing for outsiders. This is the door that keeps the taqueria honest and the banh mi five dollars: the food arrives as living practice, somebody’s Tuesday dinner, priced like a Tuesday. The second is the prestige door: the food enters top-down, through culinary schools and white tablecloths, carried by people wealthy enough to import it, already coded as refinement and already dear. The third is the pastiche door, and pastiche is the word that explains Mon Ami Gabi. Pastiche is imitation that copies the surface of a thing and leaves the substance behind: the signifiers without the soul. The pastiche channel is French food arriving not as French people cooking, but as Americans assembling the signs of Frenchness into a costume (the bentwood chair, the chalkboard, the trolley, the tricolour, the “ooh la la,” the accented menu) wrapped around an approximation of the plate. It imports the theatre and skips the kitchen.

Once you see the three doors, the table you climbed earlier rearranges itself, because every tier on it is just the deposit a particular door leaves behind. The temple is the prestige door: Escoffier’s apparatus, the culinary school, the white tablecloth, food that arrived already expensive and stayed that way. The chain and the costume are both the pastiche door: the industrial bistro and the grand brasserie are the same costume sewn at two scales, one stamped out a thousand times a week, one tailored in mirrors and gilt, and both sell the signs of Frenchness rather than the substance. The champion is the vernacular door: the lone cook serving the real food at the price the real food costs. The tiers were never a ladder of quality. They were a map of which door the food walked through.¹

Which is why the American board looks the way it does. French reached America through the prestige door and the pastiche door, and almost never through the vernacular one, because there was no French working class here cooking daube for itself on a Tuesday. So the country is overstocked at exactly the tiers those two doors deposit (temples, costumes, chains, the whole performing apparatus) and starved of the one the vernacular door leaves, the honest champion, which survives only in the few off-Strip rooms where someone decided to build it anyway. Americans only ever met French food in its dressed-up forms, expensive-and-real or cheap-and-costumed, and reasonably concluded that dressed-up was what French was. No conspiracy, no orchestrator: just a cuisine that came through the wrong doors and got misread by a country that only ever saw its formal wear. Terzaghi, off his Strip, is one man hand-carrying the food back through the door that was never built here, cooking what the French truly eat, for a community, at community prices.

If you still doubt the premium is coding and not cost, the proof is in the next restaurant over. Italian. Italian is structurally almost identical to French, a peasant-regional tradition under a refined high one, built on cheap cuts and few ingredients and forgiving technique. By the physics of the food, the two should have the same hollow middle, the same missing tier of good-and-cheap. But Italian doesn’t have it. The honest neighbourhood trattoria, the good cheap red-sauce place, is everywhere, exactly the tier you cannot find in French. Same root, same cuts, same method. The only thing that differs is that Italians immigrated poor and cooked for each other, so the vernacular carrier was on the boat, and Italian got coded, correctly, as peasant, while French peasant food got coded, falsely, as aristocratic. That is as close to a controlled experiment as this subject ever offers: hold the two cuisines side by side, subtract everything they share, and what’s left over, the entire price gap in the affordable tier, is status. Nothing else. You can prove it by subtraction.

Which is why French in America is the cleanest tasting-status case I know. Most cases are muddy, because real cost and status illusion are braided together. The sushi premium is truly ingredient-floored, and a skeptic can always say the expensive thing is simply better. French is the case where the food itself testifies against the price. It’s peasant food. There’s no luxury floor, no scarce-technique floor, and there’s a control group next door eating the same recipe for half the money. The frame promised aristocracy; the recipe is field food; the neighbour cuisine with the same recipe is cheap. A skeptic has nowhere to stand.

And here the complaint turns into something better. There is a fifth tier on that table that belongs to no single door, which is exactly why it’s the rare and valuable one. The proof tier (the grand room at honest prices, Zédel and the old Paris bouillons) is what happens when someone takes the vernacular door’s pricing and walks it through the pastiche door’s room on purpose: peasant food, peasant price, chandeliers included, the grandeur kept and the markup refused. It isn’t a deposit any single door leaves on its own; it’s a deliberate act, which is why it’s scarce everywhere and, in one country, missing altogether. France runs on it. The bouillons have for a century. London built it. Zédel, one room, on purpose. America never had it, because the proof tier is the vernacular door’s pricing dressed in a grand room, and America barely opened the vernacular door at all. So the empty box isn’t a grievance; it’s the opportunity, and it is undefended, because a mispricing with no author and no defender is the best kind there is. The food is cheap to source, the technique is learnable, the demand is real. I’m proof of it, and so is anyone who ever walked out of a place like Mon Ami Gabi knowing they paid a ceremony tax for a thawed plate. The competition in that box isn’t weak. It’s absent.

There’s one catch, and it’s why the box stayed empty. You cannot win it by performing harder. Out-trolleying the chain just makes you another costume. You win by refusing the performance and proving the plate, which turns the whole thing into a trust problem: how does a diner know, before sitting down, that this honest room is the real plate and not one more thawed one wearing the same chairs? Terzaghi, hero of this piece though he is, can’t be the whole answer: one honest cook off a Strip can prove the plate to the few hundred people who find him, but he cannot scale the trust. We measure performance constantly: the name, the popularity, the price, the count of votives. We have almost nothing that measures the plate. The thing the empty box is truly waiting for isn’t a better cook. It is waiting for honesty, because nowhere is the gap between what’s performed and what’s plated wider or more legible.

I sat at that table and my palate returned a verdict the performance was built to prevent, and the peasant honesty of the food corroborated me. Performance high, palate ordinary, and a century-old mistranslation is the whole reason for the spread. The food was never the problem. The food is honest. The coding lied about where it came from, and the cassoulet, to its credit, never agreed to the lie.

¹ And the costume’s power shows in who has to borrow it. When any cuisine tries to walk through the prestige door in America, it walks in wearing French: the menu dégustation, the prix fixe, the technique words on the card. Chicago’s first Michelin-starred Indian restaurant earned its star with a seven-course tasting menu whose signature dish is a foie gras samosa and whose dessert is a cardamom crème brûlée, Indian cooking admitted to the temple in French dress, down to the name over the door: Indienne. The pattern is measurable. When the Stanford linguist Dan Jurafsky read thousands of American menus, French on the page predicted a higher price for the plate. Nobody buys their way into the temple speaking Spanish or Tamil. The costume is the ticket, and the ticket is printed in French.

Subscribe now

We came here carrying the weight of everything Bangalore had shown us — the broken papaya salad, the abandoned European soul of Lupa, the honest conversation over coffee with a man who had been watching Bangalore’s food culture transform for fifteen years, the elaneer brûlée that arrived like an answer, the kari dosai that reminded us that the dish and the craftsman are always there if someone is willing to ask.

We came here, in other words, already thinking about what food is and what we are doing to it.

And then we went to an Instagram famous fine dining restaurant.

The announcement board outside excluded children and young adults. I should have read that as the first signal. Inside, the club music hit before the menu did. Tables seated close together on two levels in a small space, the noise bouncing between them. I stood there for a moment and felt the cognitive dissonance of it — this was not congruent with fine dining. Not the sound, not the space, not the energy. Fine dining asks you to slow down. This room was asking you to perform.

But Instagram had been insistent about this restaurant. The feed was full of it — dramatic plating, creative combinations, the visual grammar of fine dining rendered in warm Chennai light. A tasting menu. Amuse bouche. Innovative take on South Indian cuisine and fusion dishes. The language of culinary ambition spoken fluently in every post. We sat down. We were ready to be taken somewhere.

The first amuse bouche gave us reason to hope. It was innovative with technical mastery — precise, bright, the acidity and sweetness in conversation. A kitchen signaling capability. We settled in. We were ready to be taken somewhere.

Then the second amuse bouche arrived.

A mint sorbet. In fine dining, the sorbet is the most purposeful dish in the sequence — its entire reason for existing is restoration, the gentle reset of the palate between the complexity of what came before and what is to come. It is the dish that exists entirely in service of the other dishes. Getting it wrong is not a small error. It reveals something foundational about whether the kitchen understands what it is trying to do.

The mint was so overpowering, so unbalanced, with so little sugar or dairy to hold it in conversation with itself, that it did not cleanse the palate. It assaulted it. We reached for water. Water — to recover from the palate cleanser.

A palate cleanser that needed a palate cleanser.

In that single dish the entire argument of what is happening to food culture arrived on a cold spoon and announced itself without apology.

The meal continued its unraveling with a kind of terrible consistency.

The fusion dish collapsed at the bottom — a structural failure so fundamental it suggested nobody in that kitchen had eaten this dish attentively enough to know it needed solving before it reached a paying table. We struggled to plate them ourselves. No one noticed. No one came. In genuine fine dining the server knows the dish’s vulnerabilities before it arrives. They anticipate. They assist. The service at this restaurant had not eaten its own food.

The next dish arrived confused — caught between Chinese technique and Malabar spicing, honoring neither, paired with curd rice whose own strong flavor logic it could not accommodate. Two great culinary traditions reduced to a negotiation that satisfied neither. Not fusion. Confusion. A dish assembled for its description on a menu rather than for what happens when it meets a palate.

The meat dish had no umami. This ingredient is one of the richest umami sources in any cuisine — an ingredient chosen for its provocation value, its Instagram legibility, its signal of culinary courage. But umami is not visual. It cannot be photographed. And whoever developed this dish had not asked the most basic question a cook must ask — what does this ingredient want to become? The other meat dish, by contrast, was delightfully well done. The kitchen could cook. What it could not do was think beyond the frame of the camera.

This is not a bad restaurant trying. It is unfortunately pressured to become a content production facility that ended up serving food as a byproduct.

And it was full. Because Instagram said it should be.

Two days later we walked into Avartana.

It felt like walking into a spa. Pleasing colors, decor that created a tranquil immersion, tables spaced far enough apart to allow real conversation and real service. The cognitive congruence was immediate and complete — this room was telling you to slow down, to be present, to pay attention to what was about to arrive on your plate. No announcement boards. No club music. No performance requested.

Everything that restaurant had broken, Avartana quietly repaired.

The rasam arrived tableside — that most humble, most ancient Tamil preparation, the thin peppery broth that exists in every home and every mess room in this city, the dish so ordinary it is practically invisible — and it was made with such precision of technique, such clarity of flavor, such complete understanding of what rasam is supposed to do to a palate, that the table fell silent. Not because it was theatrical. Because it was true. The rasam didn’t become something else in Avartana’s hands. It became more completely itself.

My wife’s tempered bitter gourd arrived and stopped her mid-sentence. Bitter gourd — the most avoided vegetable in modern Indian restaurant menus because its bitterness is considered uncommercial, too challenging, too far outside the sweetness-and-umami comfort zone that delivery platforms have trained the urban palate to expect. Avartana put it on a fine dining tasting menu, tempered it with the precision of a kitchen that understood both the ingredient and the technique, and turned it into a revelation. That is culinary conviction — taking the ingredient the market has rejected and making it extraordinary. Not despite its bitterness. Through it.

The gongura fried chicken. The shrimp dumpling. The fennel pannacotta. The panacea dessert. Each dish carrying the same quality — a specific Tamil or regional Indian soul expressed through technique that honored rather than obscured it. Chandra and the staff moved through the room with the attentiveness that the failed restaurant’s hurried service had entirely abandoned — anticipating, noticing, present without being intrusive.

There was one honest disappointment. The menu has not changed since our visit in 2022. A restaurant with this level of conviction owes its evolution to itself. But that caveat noted — Avartana is what Tamil fine dining looks like when it knows exactly what it is and refuses to be otherwise.

And almost nobody on Instagram knows it exists.

In the same city. In the same week.

That restaurant — loud, cramped, hurried, Instagram saturated, full every night. A mint sorbet that assaulted the palate it was supposed to restore. A content production facility that inadvertently served food as a byproduct.

Avartana — quiet, precise, soulful, algorithmically invisible. A rasam that made a table fall silent. A bitter gourd tempered into revelation. A staff that remembered you.

If you want to understand what the placebo palate costs us — that is the cost. Not in argument. In two restaurant reservations made in the same city in the same week. One chosen by the algorithm. One chosen by someone who knew to look past it.

But Chennai did not only show me what food culture looks like when it performs without conviction, and what it looks like when conviction goes unrecognized. It showed me what it looks like when an entire cuisine quietly disappears without anyone noticing.

I am from here. This city is not an observation for me — it is a memory system. Every neighborhood, every mess room, every potti kadai corner carries something I have eaten and someone I ate it with. Which makes what is happening to Tamil food culture in Chennai not an intellectual argument but a personal reckoning.

Let me tell you about three dishes.

There was a place called Karpagambal mess — structured to recreate the old home setting, the kind of place that understood that a meal is not just food but a return to something. My friend and I used to go there in college for spicy adai with jaggery and ghee. We did not just eat. We consumed the emotion that the palate created from the dish. That is the only way I know how to say it and it is exactly what happened every time.

The adai — that thick, coarsely ground griddle cake of rice and multiple lentils, heavy with fresh coconut, requiring patience to slow-cook to a golden-brown crunch — arrived hot and fierce. The jaggery cut through the spice with a slow iron sweetness. The ghee pooled at the edges and carried everything together. My friend and I have tried to describe it to each other across decades and always failed. Some dishes refuse to be reduced to language because they are operating at a frequency that language cannot reach.

That is not a failure of words. That is the dish working exactly as it was designed — centuries of Tamil nutritional intelligence encoded so deeply into pleasure that you absorbed it without instruction. The rice and lentils together forming a complete protein that rivals meat. The spice activating digestion and metabolism. The jaggery providing iron and slow-release energy. The ghee carrying fat-soluble vitamins and satiety. We didn’t know any of this. We just knew it was extraordinary.

We still talk about that adai. We still cannot fully describe it. That shared inability — held between two people across decades — is itself a kind of monument to what a dish can do when it is completely itself.

Now ask for thavalai adai at a modern Chennai restaurant. Watch the server’s face. If they know what it is they will tell you they don’t serve it. If they don’t know what it is you will understand precisely how far the loss has already traveled.

My grandmother made aappam on specific mornings that felt like gifts. I can still see it — the floating, jiggling yellow of the egg in the middle, perfect to dip into with the crisp lace edges of the aappam. The batter fermented overnight, alive and volatile, the sourness developing slowly the way only time and coconut and the right hands can produce. She would swirl it in the appachatti — that specific curved pan that accepts no substitutes and tolerates no shortcuts — until those extraordinary lace edges formed. Then the lid. Then the steam. Then the patience.

After the crisp edges came the soft center — perfect for the coconut stew to soak into and melt with, the vegetables layered in sequence, the whole spices timed with knowledge stored in hands not written in any recipe, the final coconut milk added off heat to preserve its sweetness.

The perfect Sunday siesta prelude. From palate to soul — not metaphor, literal. That dish lives in me at a depth no restaurant has reached since.

The aappam has not disappeared entirely from Chennai. You can still find it in Kerala restaurants, in Chettinad joints, in dedicated evening street carts. But walk into any standard Udupi hotel or Saravana Bhavan style restaurant at 8pm and ask for it. Quick shake of the head. Twenty variations of dosa — yes. Aappam — no.

The reasons are entirely logical from a commercial kitchen perspective. The appachatti makes one aappam at a time. One pan, one burner, one human hand swirling at the correct angle, one lid, several minutes of steam. A flat thawa holds ten dosas simultaneously. The fermented batter spoils fast under Chennai’s heat with zero flexibility — unlike dosa batter which can be repurposed across a week from idli to dosa to uthappam to kuzhi paniyaram. The lace edge breaks if rushed. The fermentation cannot be standardized. The patience cannot be automated.

Every quality that makes the aappam extraordinary makes it commercially hostile. Its refusal to be rushed is its dignity. Its one-at-a-time pan is its philosophy made physical. The modern food economy encountered all of this and moved on to the ghee podi roast that photographs in fifteen seconds and sells at 200 rupees.

My grandmother’s aappam cannot be put on Zomato. You cannot standardize love. You cannot deliver patience in a plastic box in forty minutes. You cannot replicate the specific calibration of a hand that has swirled ten thousand aappams and knows without looking when the lace is right.

That knowledge lives in a generation of women whose grandchildren grew up ordering delivery. When those women are gone the knowledge goes with them. Not to a recipe book. Not to a restaurant menu. Gone.

I became a food thief in college. The victim — my wife’s food carrier. The crime — vatha kuzhambu. And I would do it again without hesitation.

Three things matter in a great vatha kuzhambu — the vegetables, the depth and balance of flavors, and the consistency of the gravy. The real version uses manathakkali or sundakkai — sun-dried berries, carefully foraged and dried, carrying a specific bitterness that is not a flaw but the architectural point. The sesame oil at exactly the right temperature. The mustard seeds. The tamarind soaked and extracted at the right consistency. The slow simmer that cannot be hurried because the depth of flavor lives in the time itself.

And then the magic — the rice appalam as garnish. That crunch against the slow-simmered depth of the kuzhambu, dissolving at the edges where the tamarind catches it, the texture completing what the flavor started. Together with rice and dal — the perfect balance of protein, digestive, spice, and umami. A complete nutritional system disguised as a meal that makes you want to steal.

I married into that vatha kuzhambu. Ate it at a family table across years. It has more memories in it than I can separate or name — the stolen college meal, the first time I ate it as a member of the family, every time it appeared and announced that this is what home tastes like.

The shortcut version throws raw vegetables into a basic tamarind gravy. It has the color. It has the sourness. What it does not have is time — the slow accumulation of correct decisions made by hands that have made this dish hundreds of times until the decisions are no longer decisions but knowledge stored in the fingers. The manathakkali’s bitterness transforming through the simmer into something irreplaceable. The depth that only arrives when nobody rushed it.

Worth to become a food thief. The shortcut version is not worth the risk.

Three dishes. Three generations of women whose hands carry knowledge that no commercial kitchen has replicated, no Instagram post has honored, no delivery platform has made viable, no discount has preserved.

The adai that two college friends consumed as emotion at Karpagambal mess and still cannot fully describe. The aappam that arrived as a grandmother’s gift — the floating jiggling yellow, the crisp edges, the soft center melting with the stew — the perfect Sunday siesta prelude. The vatha kuzhambu worth becoming a food thief for, stolen from a college food carrier and eaten at a family table across a lifetime since.

None of these dishes are gone. They live in homes, in the hands of women who learned from women who learned from women. But they are leaving the street. Leaving the restaurant menu. Leaving the shared public food culture that makes a city’s cuisine legible to the next generation.

When a dish retreats entirely into private homes it has begun its final journey. The next generation may taste it once or twice at a family table. But they will not know how to make it. They will not know what it should taste like when it is completely itself. They will not have the reference point that lets them recognize the shortcut version as a lesser thing.

And when that reference point is gone — the dish is gone. Even if the name survives on a menu somewhere, printed next to a photograph of something that has forgotten what it was supposed to be.

Mor kali disappearing because it cannot justify a premium price point. Thavalai adai vanishing because it cannot be flipped in seconds on a commercial thawa. Puli mandi and vendakkai mandi — that masterclass in zero-waste cooking built from arisi kalani, the starchy water from washing rice — gone because humble liquids cannot command high margins. Vatha kuzhambu shortcut into a basic tamarind gravy because the manathakkali requires careful foraging and the slow simmer requires time nobody will pay for. Ragi simili urundai — that powerhouse of iron and calcium shaped into a dark ball — losing to colored pastries and cheese peri-peri fries because it cannot compete visually. Elandha vadai, the Indian jujube street snack, now only in plastic packets at highway souvenir shops.

A cuisine being dismantled not by force but by the accumulated daily preference for what photographs well, delivers fast, and justifies a premium price.

Chennai knows what it is losing. It just hasn’t decided yet whether it cares enough to stop it.

And in a quiet restaurant that the algorithm has largely ignored, a chef is preparing rasam tableside and tempering bitter gourd into revelation for a room that is never as full as it deserves to be.

The answer is already here. It is just not trending.

Part 3b — The Reckoning — continues next.

Subscribe now