I know that our current Western concepts of urban layout probably don't entirely map to Ancient Rome

It's more than that, actually. It's not simply that modern urban planners lay out cities differently today, it's that cities themselves are completely different entities in the modern world. Modern cities have the benefits of rapid transit, either by privately-owned vehicles or public transport. They also are fueled by industry, which given the more or less parasitic nature of cities is significant--modern cities are able to contribute to production in a way that ancient cities could not. Modern cities are freed from walls and from the height restrictions (though Roman residential buildings could get quite tall) of ancient builders. All this means that urban structure of the modern world can give no indication of the urban structure of the ancient world. This might be obvious, but it's not necessarily so obvious in practice.

There are several models of modern urban development going back to the early part of the twentieth century. Burgess' 1925 model proposed a fairly simple set of concentric rings, with a central business area in the center, surrounded by rings of residential districts getting wealthier as they moved out, with areas of transition in between, and eventually a ring of industry at the periphery. Something like two decades later Hoyt remodeled this according to sectors, or wedges cutting through the rings along transport routes. By this model the highest-income residences mainly lie along one or two thoroughfares, often at some distance to the city, and industry and lower-income residences are close to each other. There are other models: Harris-Ullman, for example, proposes urban land use clustering at several points in a city around important areas (the central business district, or a factory complex, etc.), at which you might find smaller versions of class divides and so forth. The point is that these models all show some of the fundamental differences between modern and urban land use. Most immediately notable is probably transport. Transport at Rome, and most other ancient cities, was by foot. At the most basic level, this explains why the urban area of Rome was so relatively small despite its booming population: the city could not reasonably outgrow the space in which a human being could walk over the course of about a day. At another, it effectively limits the distance at which wealthy residences can be built and still allow reasonable access to the urban center. At Rome this is a serious problem. While senators and other magnates might--and always did--have extensive landholdings in the country, when at Rome they needed to be able to get to the forum daily. As a result you'll find the Palatine, which overlooks the forum, is prime real estate.

However, this should not necessarily mean that areas like the Palatine should be considered the Park Avenue of Rome. Neighborhoods like the Palatine, the Subura, or the Esquiline may have had their own little reputations, but in the broad scheme of things there's very little to suggest that one area of the city differed greatly from another at all. Mignone's recent book The Republican Aventine and Rome's Social Order sets out, in the first place, to challenge the idea (going back to about 1907) of a "plebeian Aventine" and in the second to show that there's not really any evidence for social stratification of this kind. The first point is not really our concern. Let it suffice to say that Mignone goes through the evidence with impressive thoroughness and shows pretty conclusively that there was nothing "plebeian" about the Aventine--in fact, we arguably have more evidence for patrician and senatorial residence there! The second point is what we want. Mignone assembles a lot of evidence, including the Regionary Catalogs and the Forma Urbis. Both are problematic--the Regionary Catalogs are almost certainly corrupt, and the Forma Urbis is highly fragrmentary--but both seem to show pretty conclusively that the large domus of wealthier residents and the clustered insulae of the ordinary worker are spread pretty evenly through the city, with no region having a particularly larger share of domus or insulae than any other. Of course, neither the Residency Catalogs nor the Forma Urbis likely show the ramshackle housing that was likely characteristic of much of Rome, but the evidence agrees with our literary evidence, which attests strongly to the close physical contact the urban elite had with the rest of the population.

Mignone goes further and, in what I think is rather a poor assertion, postulates nucleated micro-districts surrounding individual domus. The model relies too heavily on a strict model of clientship, but in any case the evidence we have does not seem to suggest that premodern Rome really had socioeconomic districts such as we might recognize them. On another level, it seems hard to see how this could be possible. Mignone's analysis more or less assumes a stable population for Rome, with permanent residents living in the same place over the course of years. In fact--and this actually supports her argument--our evidence strongly suggests that Rome's population was highly fluid. Mortality in the city to disease was extremely high, and even by the most conservative estimates immigration from Italy and eventually the provinces was necessary at a prodigious rate to keep the population stable. The population of Rome grew over time. Our evidence suggests that most of these were free people, not freedmen--the numbers of freedmen are too small. The evidence, going back to studies by Brunt in the 1960s, strongly suggests a fairly mobile urban proletariat. Rome's inhabitants were, in large part, semi-migratory day-workers, who roamed the city but also seem to have lacked permanent residences over long periods of time. They might live in the same place for a season if they found work at, say, a monumental building project, but they'd likely return to the country as hired labor at the harvest (a model that persisted until really quite recently even in the developed world) and when they returned they'd be working elsewhere. In such a fluid population how can we reasonably expect distinct districts? These people went where the work was, and since work like building projects was constantly moving (and quite frequently such monuments were not being set up in the boonies) we should expect to see labor move. Even in areas like the Tiber docks, near the Aventine, we seem to see workers coming in and out, and both rich and poor are forced to live within walking distance of the area. This is most pertinent to the question of ethnic districts, which I addressed here. But it also makes us wonder again about Mignone's micro-nucleated model. The permanent inhabitants of areas of the city were, by and large, the wealthy and the skilled workers. Skilled workers seem to make more sense as the central units of micro-neighborhoods. Their shops and workshops provided food and the necessary goods of life, and it's abundantly clear in our texts that their patronage was both impoverished and wealthy. Under Augustus these skilled workers, as representatives of their vici, became literally the cult overseers of such micro-neighborhoods. We don't see much evidence for large stratification there. Shopkeepers of the same type in particular areas frequently formed collegia, but shops appear clustered together--there is no shoemakers' district or butchers' district identifiable at Rome. The necessities of life would need, by obvious logic, to be clustered together within easy reach, and we should see this pattern repeat itself throughout the districts of the city, as these nuclei should crop up everywhere people are trying to survive. I do not believe anybody has yet tried to do that--the project would be immense, and necessarily unsatisfying. But if we could identify patterns of access to such necessities then we might be able to tease out more carefully the horizontal stratification within mico-regions, determining the relationship more closely between the large domus of a city block and the shop and tenements surrounding it