The Four Romans: One Technique, Four Dishes

The four great pastas of Rome — cacio e pepe, gricia, carbonara and amatriciana — are not four dishes that happen to resemble one another. They form a single family, and the genealogy is by now standard language in Roman food writing. Cacio e pepe is the stem. Add guanciale and it becomes gricia. Add an egg yolk to gricia and it becomes carbonara. Add tomato in place of the yolk and it becomes amatriciana. Katie Parla, the American-Italian expert who has been writing about Rome for over fifteen years, notes in a recent interview that this quartet framing is itself relatively young: ten years ago, almost no one recognised the four as a group; now every Roman trattoria does. The historian Luca Cesari has thoroughly documented the recent origin of carbonara, with the first printed recipes appearing only after the Second World War and the canonical version as we know it in fact a codification from the 1990s. Alberto Grandi, professor of economic history at Parma, situates the four within what he, following Hobsbawm, calls invented tradition: cultural authenticity as construction rather than continuity.

The genealogy, then, is well established. What is done less often, systematically, is to ask why these four dishes can be one family in the first place, and why precisely these four and not four others. The answer does not lie in the ingredients. The answer lies in a single technical problem for which all four are variations of a solution: emulsion stability.

One Problem, Four Solutions

A pasta sauce is, chemically speaking, often an emulsion — a fine dispersion of fat droplets in a water phase. Water and fat do not mix spontaneously. A third substance is needed, one that positions itself at the interface, the so-called emulsifier, which stabilises the fat droplets and prevents them from coalescing. In the Roman kitchen, starch from the pasta water does this work. On heating, starch forms a viscoelastic network that holds the fat and protein droplets in suspension. The machinery with which we can decompose processes of this kind — from starch gel to casein aggregation to lecithin as an emulsifier — is largely McGee material from On Food and Cooking (1984, revised 2004), applied here to a family he himself does not treat as a whole.

The specific physics of cacio e pepe was published in April 2025 in Physics of Fluids by Bartolucci, Busiello, Di Terlizzi and colleagues from institutions including the Max Planck Institute in Dresden, in a paper with the appropriately dry title Phase behaviour of Cacio e Pepe sauce. Below a certain starch concentration the cheese clumps; above another, the sauce becomes unworkable. There is a phase diagram with a narrow operable region.

Dario Bressanini, chemist and Italy’s most widely read scientific food writer, had already explained this on his blog Scienza in cucina in 2008: the ideal working temperature for pecorino sits around 55 to 60 degrees Celsius. Below that, the cheese melts incompletely; above it, the casein proteins denature and aggregate into stringy clumps. Bressanini makes an important observation in passing: he finds scientific cacio e pepe harder than scientific carbonara. The same, without the scientific vocabulary, appears in the well-known video by Luciano Monosilio (2021), where he explicitly calls his originale the most difficult of the three versions he presents. And J. Kenji López-Alt wrote on his Patreon in May 2026 that the four Roman pasta sauces “are all simple emulsions of fat and water with varying amounts and kinds of protein in the mix”, recommending gricia as the pedagogical starting point because it is “more forgiving” than cacio e pepe.

These three observations are connected in a way that is rarely made explicit. Cacio e pepe is the hardest because cheese is the only emulsifying medium and the temperature margin is small. Everything you subsequently add across the family is a technical softening of that problem.

What Each Ingredient Solves

Guanciale adds rendered pork fat. Fat is an emulsion donor that shifts the protein-water ratios and widens the thermal range. Where cacio e pepe demands work within a five-degree window, gricia gives you fifteen. This makes gricia not only easier, but also scalable for restaurant service — which explains why gricia flourished in urban trattorias while cacio e pepe remained a testing ground for a cook’s skill.

Egg yolk adds lecithin, a phospholipid that is among the most powerful natural emulsifiers and the basis of the entire mayonnaise industry. With lecithin, the emulsion problem becomes in principle trivial — provided the temperature stays below the coagulation threshold of egg protein, roughly 65 degrees for yolk. Hence Monosilio’s bagno-maria, which holds the mixture at precisely 70 degrees, just under coagulation but warm enough to bind. Carbonara is technically the richest emulsion of the four and simultaneously the most unforgiving at the upper end: a single moment too hot and you have scrambled eggs.

Tomato adds acid and water to a system that was built on fat and protein. This seems counterintuitive, since acid denatures protein and water dilutes the emulsion. But tomato also brings pectin and glutamic acid. Reducing tomatoes with guanciale fat for fifteen to twenty minutes builds a self-sufficient emulsion on pectin and rendered fat, one in which the pecorino at the end is almost no longer needed as a binder. Hence in the canonical amatriciana the cheese is grated over the plate, not stirred into the pan. Technically, the sauce is no longer a cheese-water emulsion but a pectin-fat emulsion. Amatriciana is therefore not gricia with tomato; it is a fundamentally different emulsion that inherits from gricia only the fat and the flavour base.

The Axis Through the Four

What you get, when you order the four this way, is a spectrum of emulsion stability that rises from left to right. Cacio e pepe sits at the hard end with cheese as the sole emulsifier and a five-degree window. Gricia widens that with fat. Carbonara displaces the problem into temperature management but delivers a structurally stable emulsion so long as you stay below the coagulation threshold. Amatriciana largely replaces the problem by building a different emulsion. Each step adds an ingredient that solves a specific emulsion problem.

This is a pedagogical ordering, not a historical one. Historically, gricia is probably older than cacio e pepe, because guanciale was self-evident in mountain shepherd kitchens. Amatriciana comes from the nineteenth century. Carbonara is a postwar American-Italian hybrid. Cacio e pepe as an urban trattoria dish with refined emulsion technique is younger than it seems. But chemically, as an analytical ordering by emulsion principle, the progression holds.

The Outsider

There is a fifth dish that is often mentioned in the same breath: spaghetti aglio, olio e peperoncino. It shares the technique — starchy pasta water binds olive oil into a crema, the same emulsion logic. But aglio e olio is Neapolitan, not Roman, and uses vegetable fat and no dairy or meat whatsoever. It falls outside the family because the family is not defined by technique alone, but by a specific combination of technique and pantry. The pantry is pecorino, guanciale, black pepper, later tomato and egg. Aglio e olio shares the mantecatura, but not the ingredient base. The dish is a good example of how cuisines differentiate themselves: the same technical knowledge spreads freely, but the canon is carried by local raw materials. Bartolucci et al., in an aside in their paper, note precisely this point — that the starch-emulsion technique also explains aglio e olio but does not make it a family member.

What the Four Preserve

If you understand the four as one technical tradition in four ingredient variants, that explains why precisely these pastas have anchored themselves with such force. Cucina povera dishes are numerous in Italy. Few survive industrialisation and the postwar internal migration as intact as these four. The reason is not primarily flavour, or not only flavour. The reason is that each member of the family is a technical learning line that supports the next. Whoever masters gricia can also make cacio e pepe and carbonara. Whoever masters carbonara can make amatriciana. The four constitute a curriculum. A kitchen that maintains this curriculum maintains not four recipes but a single craft discipline, and that discipline is transferable to sauces beyond the pasta family. Kenji López-Alt makes this point explicitly: whoever masters the mantecatura can make glossy pan sauces for chicken, fish, vegetables — anything that needs fat and water to bind.

This is the pattern one encounters more often within Nourishment, and it forms the basis of why tradition is worth maintaining. Copying recipes yields four separately executable dishes. Understanding the family yields a technique that produces four dishes and with which you can build further. That is the difference between folkloric reproduction and transferable craft. The Romans have, without formulating it as such, preserved a system in which tradition takes the form of a learning curriculum rather than a recipe catalogue. That is precisely why it survived.

---

References

Methodological: Harold McGee, On Food and Cooking: The Science and Lore of the Kitchen (Scribner, 1984; revised 2004). The chapters on starch-thickened sauces, emulsions, casein behaviour and egg coagulation supply the chemical machinery on which this essay leans.

Historical: Ada Boni, La cucina romana (1930), for the canonical Roman repertoire before carbonara. Luca Cesari, Storia della pasta in dieci piatti (Il Saggiatore, 2021; English: A Brief History of Pasta, 2023), for the historical demythologisation. Alberto Grandi, Denominazione di origine inventata (2018) and La cucina italiana non esiste (2023, with Daniele Soffiati), for the invented-tradition frame.

Culinary journalism: Katie Parla, Tasting Rome (2016, with Kristina Gill) and her Saveur column, for the popularisation of the four-family framing.

Scientific: Dario Bressanini, Scienza in cucina blog (2008 onwards) and his Gambero Rosso interviews, for the physical-chemical analysis of cacio e pepe. G. Bartolucci et al., “Phase behaviour of Cacio e Pepe sauce”, Physics of Fluids 37:4 (April 2025), for the experimental phase diagram.

Chef perspective: J. Kenji López-Alt, “Master this Technique and You’ve Mastered Roman Pasta: Pasta alla Gricia” (Patreon, May 2026), for the explicit emulsion framing and the gricia pedagogy. Luciano Monosilio, Cacio e pepe: originale vs. infallibile vs. gourmet on Italia Squisita (2021), for the chef’s typology of the same sauce in three difficulty levels.