Is Mersu the Oldest Dessert? Why the Claim Makes No Sense

Pop history creators, whether videos or text, love to reduce complex food history questions to easily digestible factoids. One such sensational claim concerns the origin of dessert. According to this claim, mersu, a 4,000-year-old Babylonian preparation of dates, flour, and fat, is “the world’s oldest dessert.” Certainly, this makes for an irresistible hook, promising a direct culinary link between our modern sweet treats and the dawn of civilization. Yet while cuneiform tablets confirm that mersu is indeed one of humanity’s oldest recorded sweet recipes, calling it a dessert is flatly wrong.

The concept of dessert did not exist until meals were separated into distinct temporal stages.

The error isn’t in the archaeological record, but in how we read it. A food cannot belong to a dining category that did not exist when it was eaten. In ancient Mesopotamia, there was no concept of a sequential multi-course meal, no ritual of clearing the table to serve a new course, and no actual category called “dessert.” Crowning mersu as the first dessert is an act of historical back-formation, taking an ancient, sacred food preparation and shoving it into a modern cultural category where it does not belong.

What Was Mersu? (The Real Babylonian Recipe)

What we know about mersu we learned from the cuneiform tablets excavated from ancient Mesopotamian sites, most notably the 18th-century BCE royal archives of Mari in modern-day Syria. In these administrative texts, the word appears as mersu (or mērsum), derived from the Akkadian verb marāsu, which literally means “to stir,” “to blend,” or “to mash together.”

The name described the food’s physical preparation rather than a finished bakery item. Mersu was essentially a dense, rich confection created by pounding dried dates into a paste and thoroughly working in flour, clarified butter or animal fat (samnu), and often water or oil. Depending on the occasion and the wealth of the kitchen, high-status variations called for shelled pistachios, figs, apples, or crushed coriander and cumin seeds.

Because ancient scribes rarely recorded step-by-step baking instructions, assuming the palace cooks already knew their craft, modern culinary historians have long debated its final texture:

Some propose a thick, sweetened flatbread or leavened cake baked on hot stones or clay platters.
Others suggest a raw, pliable dough shaped into rolls or balls, functioning much like modern date-nut energy truffles.
A few interpret it as a rich, dense porridge or pudding paste stirred over heat.

Since we have no way to determine the ingredient proportions, we can’t be sure of the final form or texture of the food. However, whatever its physical form, mersu was not an everyday peasant staple or a casual mid-afternoon snack. It was an elite food item filled with luxury resources like preserved sugars and calorie-dense fats. More importantly, its presence in administrative records reveals that it was overwhelmingly prepared as a holy food for temple rituals and royal banquets. Priests offered mersu directly to the deities, most notably during the sacred rites for the goddess Ishtar, serving as divine nourishment before high-ranking officials and priests consumed the consecrated leftovers. It was treated with holy reverence and political prestige, not as a casual sweet treat to cap off dinner. And, regardless, such a concept of a food specifically eaten after a “savory meal”, to be abundantly clear, did not exist.

The Linguistic Rule: Why “Dessert” Requires an Empty Table

It may be difficult to understand why claiming the ancient Mesopotamian ancient culture invented dessert is so fundamentally flawed. The key is in understanding what the word dessert actually means. In popular usage, “dessert” is often treated as an all-purpose synonym for anything sweet. However, if you grab a chocolate bar from a vending machine in the middle of the afternoon, you might think of it as a sweet treat or a snack, but it’s not a dessert.

In other words, dessert is not defined by its ingredients, its sweetness, or its sugar content. It is defined strictly by its temporal position in a meal.

Our English word comes directly from the French verb desservir, which literally translates to “un-serve” or “to clear the table”. The term traces back to the Latin servus (servant or slave) and servire (“to serve”), which combined with the negating French prefix des- to mean the literal act of removing what had previously been brought to the table.

Historically, the word did not initially refer to food at all. It referred to the physical act of changing the dining space, consisting of the clearing away of dirty trenchers, meat platters, sauces, breadcrumbs, and salt cellars once the main savory meal had concluded. Only after the table was physically “un-served” did hostesses present a final, ceremonial offering of sweet wines, candied fruits, or nuts. Over time, the name for the action of clearing the table transferred onto the sweet items consumed immediately following that clearance. By 1539, records show the term was firmly established in France to describe this concluding, post-service ritual.

This distinction exposes the impossibility of an ancient “dessert.” The entire concept requires a rigid, sequential dining architecture:

A meal must be divided into distinct, successive courses.
Savory dishes must precede sweet ones.
The previous savory course must be physically removed from the dining surface before the sweet course can be presented.

Ancient Mesopotamian and Mediterranean dining had no such architecture. Banquets and everyday meals in the ancient Near East did not progress through sequential, compartmentalized courses. Food was served concurrently and eaten as communal spreads. Sweet elements like honey, date syrup, fruit pastes, and stewed fruits, were eaten alongside roasted meats, lentils, and bread, or incorporated directly into the main savory dishes themselves. There simply was no dessert and sweet foods did not carry a particular connotation as being separate from other foods.

This type of back-formation is not isolated in food history circles. Claiming mersu is the “earliest dessert” relies on the exact same error explored in The Myth of Ancient Fast Food, where casual trivia features look at ancient Roman tavern counters and label them the “world’s first McDonald’s” simply because they handed hot lentils to a worker. It is an act of linguistic and structural back-formation. Just as a Roman street stall lacked the industrial assembly lines, centralized supply chains, and franchise logistics required to be “fast food”, an ancient Mesopotamian temple dish lacked the sequential meal structure and table-clearing ritual required to be a “dessert.”

Why You Get Your “Just Deserts,” Not Your Dessert: The phrase “to get one’s just deserts” frequently confuses modern readers, who assume it references an after-dinner treat or an arid wasteland. It is neither. While the word is spelled like the barren landscape, its pronunciation and ancestry tie directly back to the Latin root of dessert.

Both words originate from the Latin servire (“to serve”), but they split along different French branches. While desservir used the prefix des- as a negator to mean “un-serve” (clear the table), Old French also developed deservir, where the prefix acted as an intensifier meaning “to serve fully” or “to merit.” By the late 13th century, this entered Middle English as the noun desert, meaning a worthy reward, merit, or punishment according to what one has earned. When someone receives their “just deserts,” they aren’t being given a slice of pie. They’re receiving what their actions fully served up.

The Category “Dessert” Came Centuries After the Food

Once you understand that dessert is defined by a cleared table and a specific dining sequence, the premise of searching for humanity’s “first dessert” becomes a fool’s errand. It treats dessert as though it were a tangible, discrete invention, like the wheel, brewing, or leavened bread, rather than what it actually is, a late-emerging cultural category.

When French banquet culture began formally staging the dessert course in the 16th and 17th centuries, diners didn’t suddenly discover the concept of sweet foods. Instead, an aristocratic dining tradition developed a strict ceremonial protocol (service à la française) that reorganized a vast, pre-existing array of confections that Europeans and Mediterranean cultures had already been eating for centuries.

As linguist Dan Jurafsky points out in The Language of Food, early French dessert courses consisted of items that were already ancient culinary staples:

Spiced wines like hippocras
Fresh and dried fruits, such as figs and raisins
Preserved fruit comfits and seed dragées (nuts and seeds coated in boiled sugar)
Aged cheeses and crisp wafers

None of these individual foods became desserts until the culture created a temporal and categorical slot for them to occupy. A handful of candied seeds or dried dates eaten on their own was simply food; provisions consumed for sweetness, sustenance, or pleasure without any formal dining classification. They only transformed into ‘dessert’ when placed on a table that had just been un-served of its roast meats, broths, and trenchers

There was never an ancient baker who created the inaugural dessert, because the category itself did not exist until thousands of years after those recipes were written down. Looking for the “oldest dessert in history” in 2000 BCE Babylonia is like scouring ancient Pompeii for the “world’s first sports car” simply because Roman chariots had wheels and moved quickly down paved roads. It mistakes a shared ingredient, sugar or speed, for an entire economic, social, and structural system that belonged to a completely different era.