> ## Content Index > Fetch the complete content index at: https://romanempiretimes.com/llms.txt > Use this file to discover other available public pages before exploring further. # The Surprising Roman Roots of Macaroni and Cheese - URL: https://romanempiretimes.com/surprising-roman-roots-of-macaroni-and-cheese/ - Published: 2026-10-04T00:41:02.000Z - Updated: 2026-10-04T00:41:02.000Z - Description: Cato layered grain dough with fresh cheese centuries before macaroni and cheese. His Roman recipes reveal a surprising backstory, with important differences. - Author: George Liapis - Tags: Roman Empire Anecdotes, Roman food, food, Cato the Elder, Greek, olive oil, apicius A Roman recipe written more than two thousand years before today’s macaroni and cheese calls for sheets of dough layered with fresh cheese, enclosed in a crust and baked. It also calls for honey. The dish is *placenta*, recorded by Cato the Elder in *On Agriculture* around the middle of the second century BCE. Its ingredients make the modern comparison tempting; its method shows why the comparison still needs verification. A 2026 Columbia University Press history of macaroni and cheese begins with this Roman preparation. Its argument is about an ancient antecedent, not a claim that Cato made the macaroni and cheese served today. Cato does not instruct anyone to boil macaroni or prepare a cheese sauce. He describes a substantial baked assembly of grain sheets and cheese, to many a Roman early version of the mac and cheese. The Roman evidence becomes more surprising when Cato’s next recipe leaves out the honey, while a later Roman collection cooks broken pieces of dried dough in milk but adds no cheese. Taken together, the texts preserve several parts of a familiar food idea that after centuries of human history, could have led to the finished modern dish. Following those parts through the surviving recipes gives mac and cheese a backstory that begins much earlier than its name. ## A Roman estate manual preserves the first clue Cato wrote *On Agriculture* for the management of a Roman estate. His instructions range across farm work, equipment and food preparation. The book reads the estate manual in the context of Cato’s attachment to older Roman practices and the sizeable estate household that could make use of its recipes. As with all cookbooks, the existence of an instruction in such a text tells us what its author thought worth recording; it does not establish that everyone in Rome ate the result. The name *placenta* comes from a Greek word associated with something flattened, as the book explains. In Cato’s recipe, the cook first makes *tracta*: sheets formed from grain dough, rubbed with oil and left to dry. Cato says to soften groats, work them with flour and water, shape the dough on an oiled board, and transfer the sheets to drying baskets. The instructions distinguish preparation of the sheets from preparation of the enclosing crust; a separate dough provides that crust. These are not the small curved pasta shapes implied by the modern word “macaroni”; Cato specifies broad sheets and a layered construction, somehow a blend between lasagna and the modern version. Cato’s instructions then turn to fresh sheep’s cheese. The cook soaks and drains it, works it smooth, and mixes it with honey. Sheets of *tracta* alternate with the cheese mixture until the filling is used. The assembled food sits over bay leaves, is wrapped in its crust and bakes beneath a heated cover. The ancient text documents the sequence of preparation, while the modern study explains why a modern kitchen reconstruction necessarily adapts the equipment and ingredients. ## The cheese layers were baked, not boiled The scale alone separates Cato’s recipe from a quick bowl of pasta. His full version specifies fourteen pounds of cheese and four and a half pounds of honey. The book interprets this as food for a gathering, associated with celebration and the means to use an expensive sweetener. There is a more basic culinary difference. In the version preserved by Cato, the dough sheets are assembled dry with the moist filling and baked inside another dough layer. The recipe does not first boil the sheets, melt hard grated cheese over them, or make a sauce. The book notes a disagreement among scholars about whether *tracta* in this setting should count as pasta at all. That uncertainty does not make the ancient recipe irrelevant. A surviving Roman writer gives unusually concrete directions for putting grain sheets and cheese together, and the dish’s structure can be compared with later preparations. The feature photograph of modern shell pasta illustrates today’s macaroni and cheese, not what *placenta* looked or tasted like. ![Open fifteenth-century manuscript containing Cato’s De agri cultura and Varro’s De re rustica in the Biblioteca Medicea Laurenziana](https://storage.ghost.io/c/aa/80/aa80281a-abce-4f35-865f-05153c8cc335/content/images/2026/10/cato-de-agri-cultura-fifteenth-century-manuscript.jpg) Fifteenth-century manuscript containing Cato’s **De agri cultura* and Varro’s **De re rustica*, Biblioteca Medicea Laurenziana. Credits: [Sailko / Wikimedia Commons](https://commons.wikimedia.org/wiki/File:Italia,%5Fm.p.%5Fcatone,%5Fde%5Fagri%5Fcultura,%5Fe%5Fvarrone,%5Fde%5Frustica,%5FXV%5Fsec.,%5Fpluteo%5F51.2.JPG?ref=romanempiretimes.com) ([CC BY 3.0](https://creativecommons.org/licenses/by/3.0/?ref=romanempiretimes.com)) ## Cato’s second recipe removes the honey Immediately after *placenta*, Cato records *scriblita*. Its short instruction refers the cook back to the earlier recipe’s crust, *tracta* and cheese, then changes one ingredient. In the 1934 translation by W. D. Hooper and H. B. Ash, the complete direction reads: “Follow the same directions with respect to crust, tracta, and cheese, as for the placenta, but without honey.” The omission is striking because it leaves a baked combination of dough and cheese without the ingredient that may lead a modern reader to classify *placenta* as a sweet cake. It still does not produce boiled macaroni or a cheese sauce. The book treats the variation as part of a longer search for the dish’s antecedents, while the recipes themselves set a narrower limit: they establish ingredients and method, not a continuous descent to the present. Cato also records *libum*, a cheese-and-flour loaf, in the same group of preparations. These different instructions show that cheese could be combined with grain products in more than one form. Unfortunately, they do not offer us details, like how frequently each food appeared on Roman tables. 🏛️ Our account of [recipes preserved from ancient Rome](https://romanempiretimes.com/some-recipes-date-back-to-ancient-rome-french-toast-foie-gras-and-braised-flamingo/) offers other examples of the gap between a surviving text and the full range of ancient eating. ## Another Roman text supplies the milk, but not the cheese The book draws attention to a preparation in the collection conventionally called *Apicius*, compiled centuries after Cato. The recipe for *pultes tractogalatae* calls for dried disks of *tracta* to be broken into hot milk and water. The fragments are cooked with stirring; the directions later add salt and oil. Cheese is absent. The preparation therefore does not complete a Roman macaroni-and-cheese recipe, even though it supplies a different way of cooking grain pieces in a dairy liquid. The word *tracta* also presents a problem of interpretation. The book argues that its form and use can change across Roman texts. Some scholars read the broken sheets in *Apicius* chiefly as a thickener; it suggests they may also have provided edible pieces, with their starch contributing to the liquid. The terse instruction does not say whether the pieces should dissolve or remain distinct. It tells the cook to stir to prevent sticking, but gives no cue for the final texture, something that is allowing us to leave it open to interpretation. The two Roman sources consequently offer an intriguing reversal. Cato has dough sheets with cheese but bakes them; the later milk preparation cooks broken dough pieces in liquid but has no cheese. The contrast is more informative than the simple assertion that Romans “invented mac and cheese,” but it clearly shows how early versions of pasta and cheese or milk combinations, were possibly frequently enjoyed during the Empire. ## How the familiar combination acquired a new form The book "*The Epic History of Macaroni and Cheese*" by Karima Moyer-Nocchi follows the story beyond Rome to medieval recipes in which boiled dough and cheese appear together. It warns that “macaroni” did not always mean today’s elbow-shaped pasta: the word could cover several forms of dough. A Genoese inventory from 1279 records a basket of macaroni, but gives no recipe. The English *Forme of Cury*, compiled around 1390, gives a clearer recipe called *makerouns* or *makrows*. It tells the cook to cut a thin sheet of dough, boil the pieces, and layer them with grated cheese and butter. This later text supplies boiling and the butter-and-cheese finish missing from Cato. The book places it among other medieval Italian and English recipes that adapted the pairing in different ways. It does not present one unbroken recipe copied unchanged from a Roman estate to an English kitchen; rather, its evolution. The wider history follows further changes across social settings and centuries. The Roman chapter remains the arresting opening because a named ancient writer left detailed instructions for layering grain dough with cheese. The medieval evidence shows why that ancient starting point should be described as a backstory rather than an invention date. Readers interested in the practical setting of Roman cooking can also see our report on [the excavated Roman kitchen at Segisamo](https://romanempiretimes.com/roman-kitchen-segisamo-sasamon-spain/). ## What the Roman roots really tell us Ancient recipes are unusually direct witnesses to combinations of ingredients and sequences of work. They are less direct witnesses to taste, ordinary frequency or the paths by which a preparation moved between communities. Cato’s estate manual does not establish a universal Roman menu, and the much later manuscript photograph cannot close that gap. Our recent account of [how Romans preserved food without refrigeration](https://romanempiretimes.com/no-fridge-no-problem-how-romans-preserved-food/) provides further context for food practices. The Roman evidence invites a specific conclusion. Cato’s *placenta* layered dough and cheese with honey; *scriblita* kept the layers and dropped the honey; a later Roman text cooked broken *tracta* in milk without cheese. Each recipe resembles one aspect of a later comfort food that most enjoy. The surprise lies in how much of the pairing is documented early, and how much changed before boiled macaroni, cheese and butter appeared together in medieval writing. The recent Smithsonian’s [Cooking Up History event on the global journey of macaroni and cheese](https://americanhistory.si.edu/events/calendar/cooking-history-global-journey-macaroni-and-cheese-karima-moyer-nocchi-?ref=romanempiretimes.com) offered one contemporary occasion to discuss that longer history. For the Roman chapter, the firmer starting place is still the recipe text: an ancient cook was told to build layers of grain dough and fresh cheese, and to bake them. Perhaps one Roman's inspiration, led millennia later to our modern mac and cheese. 🏛️ ****Sources used:** Karima Moyer-Nocchi, **The Epic History of Macaroni and Cheese: From Ancient Rome to Modern America* (Columbia University Press, 2026), Introduction and Chapter 1, supplied ebook; [publisher record](https://cup.columbia.edu/book/the-epic-history-of-macaroni-and-cheese/9780231215312/?ref=romanempiretimes.com). Cato the Elder, **On Agriculture* 74–78, translated by W. D. Hooper and H. B. Ash (1934): [ancient text and translation](https://penelope.uchicago.edu/Thayer/E/Roman/Texts/Cato/De%5FAgricultura/E%2A.html?ref=romanempiretimes.com). Smithsonian National Museum of American History, [Cooking Up History event page](https://americanhistory.si.edu/events/calendar/cooking-history-global-journey-macaroni-and-cheese-karima-moyer-nocchi-?ref=romanempiretimes.com).