Stuart Franklin/Magnum Photos

The Roman Forum in 2008, looking east toward the Arch of Titus, with the remains of the Temple of Saturn (right), the Temple of Castor and Pollux (center, with three columns), and the Temple of Vespasian (foreground); the Palatine Hill rises to the right

The ancient Roman Forum is one of the most frustrating tourist sites in the world. This is the spot where some of the most famous events in Western history took place and some of the most consequential decisions were made. It is where the Roman Senate debated how to respond to the threat of Hannibal, where Marcus Tullius Cicero denounced would-be tyrants and radical revolutionaries, where Julius Caesar’s body was cremated after his assassination in 44 BC, and where Mark Antony delivered the original version of “Friends, Romans, countrymen.” Yet what you now see has almost nothing to do with any of that.

The imposing “senate house,” preserved to more or less its full height thanks to its conversion into a church in the seventh century AD, has no connection with the place in which Cicero held forth in the first century BC; it was completely rebuilt almost five hundred years later. The elegant circular temple of the goddess Vesta (where the Vestal Virgins kept the sacred flame of the city permanently alight) owes more to Mussolini’s “restorers” in the 1930s than to any ancient Roman builders or architects. The ground surface is largely a confusing mass of rubble and masonry, interspersed with equally confusing holes left by archaeologists digging down in search of the structures, shrines, and burials that formed the first layers of human occupation in the city of Rome, as far back as the eighth century BC. Even the trained eye finds it hard to work out how any of this fits together, or what the place would have looked like at any particular period of antiquity. Most visitors walk through the Forum baffled. Cicero would not have recognized it.

What is left of the imperial palace on the Palatine Hill, which rises above the Forum, is hardly less frustrating for the modern visitor—and endlessly debated among specialists. For centuries after the end of the Republic, this vast complex of buildings was the hub of Rome’s empire, the main residence of its emperor, and luxuriously equipped to match. Contemporary poets—whose talent for exaggeration probably did not extend to outright invention—described the precious colored marbles imported from all over the Mediterranean lining its walls, the hundreds of columns (enough to support the whole world if Atlas should decide to take a break), and the enormous height of the building, which dwarfed even the pyramids of Egypt.

Not now. All that decoration has long since been looted, leaving for the most part rough brick walls more reminiscent of a factory or warehouse than of a palace. But even more to the point than the predictable loss of splendor is the fact that, for most of us, the site has become very hard to decode. The emperors’ own enthusiasm for domestic improvements is largely to blame, as each successive modernization program covered over, adapted, or destroyed what had been there before. The result is that it is often impossible to work out, among the maze of walls, exactly what part or period of the palace we are looking at, still less what went on there.

That is true even in the better-preserved sections. One of the highlights of the Palatine, and a pleasant relief from all the bare brick, is a small series of attractively painted, but not aggressively luxurious, rooms that are said to belong to the “House of Augustus,” the first emperor. They almost certainly were part of that original “palace” (the style of the painting more or less guarantees the date). But how these rooms fitted into the overall scheme of the emperor’s accommodation is far from clear. Did they actually form the main nucleus of his living space and public reception areas, the heart of the Augustan regime? If so, that would match the claims of his biographer Suetonius that Augustus lived relatively modestly by the standards of later emperors. (And there have even been some optimistic attempts to identify a small chamber partly surviving on a floor above them with the emperor’s private study-cum-hideaway that Suetonius also mentions.)

Or were these rooms only one rather ordinary, even lowly, part of Augustus’s very much larger and more lavish palace, most of it later destroyed, its precise dimensions unknown? That is the more common recent view, despite the ancient assertions of Augustan modesty. It is rather like the problem of the emperor Nero’s “Golden House,” some of which is still preserved underground, behind the modern Colosseum metro station. There is a nagging suspicion that many of its vast eerie corridors and delicate paintings, which have enthused visitors ever since the painter Raphael explored them in the sixteenth century, had little to do directly with Nero at all, but were actually part of the service quarters.


Some of these questions of “Roman topography”—as the subspecialism of classical archaeology devoted to the layout of the city is now known—can seem narrowly arcane. The exact route of a deeply buried ancient street or the name of some minor temple may not matter very much; and pinpointing Augustus’s private hideaway, though intriguing, is in the end no more than a curiosity. But other puzzles and problems not only defeat tourists trying to make sense of these sites with their guidebooks (whose usual mode simply adds to the confusion by pretending that there are no problems at all); they also raise some important historical issues.

The size and style of Augustus’s “house” on the Palatine make a huge difference to how we understand the character of the first emperor’s rule and his relationship with the traditional Roman aristocracy whose power he had effectively displaced. Did he live in conditions that were more or less on a par with any other member of the Roman elite? Had he already sponsored a palace on a scale that far outbid the rest of his onetime peers? Or—as the fashionable view was among archaeologists thirty years ago (though almost certainly wrong)—did he live in quarters that were relatively modest in themselves but cleverly integrated into the next-door Temple of Apollo, with the obvious implications that, whatever his disdain for luxury, he really belonged very close to the realm of the gods? Equally, the arrangement of the Forum (how many people could come together there, for example, or what precisely the physical relationship was between the meeting places of the Senate and the assemblies of ordinary citizens) has enormous bearing on how we understand the working of Roman democracy during the Republic.

It is partly with those big issues in mind, and partly from antiquarian zeal, that archaeologists have tried for centuries to solve these puzzles and impose some order on the apparent chaos of the archaeological remains, by constructing atlases and models of the ancient city. The detailed three-dimensional reconstruction in plaster created by Italo Gismondi for Mussolini’s exhibition in 1937 to celebrate the two thousandth anniversary of the birth of Augustus is still the most famous such model, and continues to be reproduced on any number of posters, postcards, and fridge magnets, despite its proud Fascist origins. (It is now housed at the Museo della Civiltà Romana at EUR, though currently not on display while the museum is closed for renovation.)

Alternatively, in a very different format, they have compiled huge “topographical dictionaries” that discuss one by one, in alphabetical order, all the buildings in ancient Rome, whether they are known from physical remains, literary references, or both. The most comprehensive is the multilingual Lexicon Topographicum Urbis Romae (Topographical Dictionary of the City of Rome), edited by Eva Margareta Steinby, which appeared in six volumes between 1993 and 2000 (with five volumes following on the Suburbium, “the Suburbs”). Leaving aside the well-known public monuments, almost two hundred pages of this work, and not far short of a thousand individual entries, are devoted to private houses (from the “Domus: Marcus Acenna Cesillanus” to the “Domus: Volusius Saturninus,” both known only from their owners’ names stamped onto lead pipes). That alone gives an idea of the scale of the problems—or the richness of the data, depending on how you choose to look at it.

The Atlas of Ancient Rome, edited by Andrea Carandini, is the latest hefty addition to these topographical reference books. Now almost eighty years old, born into the liberal elite (his grandfather was director and part owner of Corriere della Sera before being forced out by the Fascists; his father was ambassador to London after World War II and president of Alitalia) and a left-wing radical (onetime member of the Italian Communist Party), Carandini is one of the most distinguished and charismatic Italian archaeologists of his generation, perhaps ever.

In many ways, he has been a tremendous force for good in archaeology and heritage. His excavations in the late 1970s of the Roman agricultural, slave-operated estate at Settefinestre north of Rome have remained a classic in the field—partly because they were concerned more with the infrastructure of Roman agriculture and social relations than with the unearthing of great works of Roman art—and recently he has been an influential president of the Fondo Ambiente Italiano (the Trust for the Italian Environment, more or less the equivalent of the British National Trust), which is concerned with the protection of Italy’s natural and artistic heritage.


Carandini’s Atlas comes in two exquisitely produced volumes, packed in a presentation box, with a ribbon thoughtfully provided to help the reader remove them; like its original Italian version, it is an extraordinarily pleasurable book to hold and handle. The first volume consists, after Carandini’s introduction, of a series of essays by different authors, many of them his colleagues and former students. These cover some general issues of Roman topography (boundaries, roads, aqueducts, and so on), and go on to cover, in individual chapters, the history and layout of each area of the city, starting with the Forum and ending with what we now know as Trastevere, on the west bank of the Tiber.

Sir John Soane’s Museum, London

Henry Parke: A Student on a Ladder Measuring a Corinthian Capital at the Temple of Jupiter Stator, Rome, 1819

The illustrations are many and often magnificent, though in a few the color is misleadingly garish. They include detailed plans and photographs of the archaeological remains, of excavations in progress, and of some key objects and works of art found in them—as well as a range of careful digital reconstructions of individual monuments and of wider areas of the city (as always there is a slightly sanitized, computer-game feel to some of these and to the toga-clad people introduced to give scale, but that no doubt comes with the genre).

The second volume contains over four hundred pages of maps, additional reconstructions, and diagrams of all sorts: from illustrations of the architectural orders, or of monuments depicted on Roman coins, to plans of different areas of the city at different periods. At first sight, it appears to be the reliable answer to anyone’s confusion about the Forum or Palatine, or indeed about any part of the ancient city of Rome. But it is not.

That is partly the fault of the translation from Italian into English, which lies somewhere on the spectrum between awkward and incompetent, and which can hardly have been attentively edited. Perhaps many readers will concentrate on the plans and illustrations, which are such an important part of the book. So much the better: because as soon as they turn to the essays, they will be as irritated, baffled, and sometimes seriously misled as those trying to make sense of the Forum on the ground. “Archaeological cartography, of the digital variety in particular, is no longer seen as an appendix to the topographical lexicon, used merely as a link between the philological-archeological data and the territory” and “a reformed cultural mindset is a precondition for carrying out an archeological endeavor” are typically awkward versions of Carandini’s flowery, though consistently lucid, prose.

But there is also a series of strange coinages such as “planimetry” for the Italian planimetria, meaning “ground plan” (“the planimetry was rebuilt based on the domus next to it” is the translator’s sorry substitute for what should be “the ground plan has been reconstructed on the basis of the house next door”); not to mention unhelpful “translations” of technical terms (the strange “hauspice/s” is used for the “auspices” or auspicia, the system of signs by which Romans ascertained the will of the gods) and frequent slips in Latin (urbis for urbs is only one example of confusion between genitive and nominative). Careless typos are suspiciously common in the names of European scholars: “Mar Augé” for “Marc Augé,” “Hesbery” for “Hesberg”; even the original publisher of the Atlas appears as “Electra,” not “Electa.”

Worse still, there are many occasions on which mistranslations or sloppy editing has produced howlers that would shock the Italian writers. The large inscribed map of the city of Rome, the so-called “Marble Plan” of which hundreds of pieces still survive, at the precise scale of 1:240, is misdated by half a millennium. It was erected in the third century AD, not BC (and most of the public “does not know” of it; they do not willfully “ignore” it as the translation puts it, missing the sense of the Italian ignora). The very idea of such a map, as well as the technical know-how to construct it, would have been unthinkable at any such early date.

At one particularly low point, the translator concocts a statue base on the Palatine that “reproduced the bas-relief of the so-called Sorrento Base, a work by Greek artists, Skopas among them, originally made in the fourth century BC.” Almost everything about this simple sentence is garbled. The original author was, in fact, clearly referring to a famous piece of relief sculpture known (from where it was found) as the “Sorrento Base,” which dates to the first century AD, not the fourth century BC. It is famous because it appears to reproduce in its design the cult statues, now lost, of Apollo, Diana, and Latona that stood in Augustus’s new Temple of Apollo on the Palatine and had been brought, so it is often believed, to Rome from Greece; it is these lost statues that were possibly the work of the fourth-century-BC sculptor Skopas. Reader, beware.

The translation is not the only weakness of the Atlas. There were even bigger problems, of audience and approach, already in the original Italian version. First of all, it is not clear at whom the book is aimed. Its lavish production suggests that it was designed as much for the coffee table as for the library shelves, as much for the interested amateur as for the specialist (and Carandini, in a new preface, refers to “the enthusiastic reception it received from the public” in Italy). But in truth, it makes hard reading for anyone who is not already reasonably familiar with Roman topography.

Even the organizing principles will be unclear to most people. Each of the main chapters or essays focuses on one of the fourteen numbered “regions” into which the emperor Augustus divided the city, for presumably administrative purposes (the area of the Forum being number VIII, the Palatine number X); and specific monuments are given an individual identifying number, or “individuated,” within each region (the Augustan Temple of Apollo, for example, being X. 13). But the nature of these regions is only very briefly explained, and it is mentioned merely in passing that we do not know their precise boundaries.

Nor is there any proper discussion of the so-called “Regionary Catalogues,” even though they are referred to frequently (and if you have never heard of them, rather mystically) throughout the Atlas. These are two lists or gazeteers, compiled in the fourth century AD, of buildings and other landmarks in Rome, ordered by the same Augustan regions. They are both an extraordinarily valuable resource for any study of the city and also in many respects very puzzling. What, for example, were they for? And how accurate is the information—including some rudimentary statistics—they contain (“130 houses, 18 warehouses, 85 bath buildings” in region VIII)? The Atlas appears to assume a readership that does not need to be introduced to the Regionaries, their background and dilemmas. Perhaps that is true of the public in Italy (though I doubt it); it is certainly not true in the Anglophone world.

It also assumes a readership that prefers dry detail to any broader overview. It is true, there are the general essays at the beginning on “the city as a whole,” and Carandini’s spirited new preface includes fighting talk on a variety of topics: on “resuscitat[ing] Rome as the ancient city was conceived,” on the way in the modern world “working-class people” have been “pushed…out of the historic city centers,” and on the importance of archaeology as a major tool for understanding the past, equal to literary sources (“history is a lady and archeology is merely her servant” is how he rather quaintly characterizes the approach he is trying to unseat). However, the main chapters on each of the fourteen regions do not do much to set the general scene. Instead they rather uncompromisingly plod through the urban development of the area, period by period, monument by monument.

If you know what information you are looking for, you will probably find it in the Atlas. But there is very little effort to explain why there are disagreements and uncertainties in making sense of the layout of the city. Nor is there much attempt to represent different views and different reconstructions. We sometimes learn that an old theory of dating or identification is now deemed incorrect, but rarely why it is wrong. In fact, the rather scant attention given to alternative interpretations or potential disagreements is bound to make the Atlas less useful even to specialists than it appears. There is, to use a grammatical analogy, far too much of the “past definite” here.

That is especially the case with the lengthy sections on the earliest history of the city, in which Carandini’s own vision of the subject is enshrined with very little indication of just how controversial it is. I heard him recently talk very powerfully at the American Academy in Rome on how archaeology should, and can, engage the public, and how important it is in that project for archaeologists to be able to tell stories about the distant past. For Carandini narrative is a fundamental part of popular, and indeed academic, archaeology.

But sometimes his stories become part of the problem. Over the last few decades he has been mostly concerned with questions of how the city of Rome began, and he has led major excavations around the Forum and the Palatine aimed at revealing the founding phases. His conclusion (as well as his premise, I suspect) is that the legends of early Rome, as told by historians such as Livy, were more or less accurate and can be authenticated by archaeology. In fact, in a series of well-publicized discoveries, Carandini boldly claims to have found such emblematic structures as the first defenses of the city built by Romulus himself, sometime very close to the traditional founding date of 753 BC; and he even thinks he has pinned down the very house occupied by Tarquinius Priscus, or Tarquin, the legendary fifth king of Rome.

Tarquin is supposed to have lived around the turn of the seventh and sixth centuries BC and is credited with all kinds of urban improvements to the proto-city, from building the “Great Drain” or Cloaca Maxima to starting the Temple of Jupiter Optimus Maximus on the Capitoline Hill. The house Carandini identifies with this legendary figure was uncovered in excavations near the Arch of Titus at one end of the Forum, though the remains are very scant (this is the house whose “planimetry was rebuilt based on the domus next to it”). Carandini has reconstructed it with a central hall or atrium, and rooms opening off—complete with the upstairs window through which, according to Livy, Tarquin’s widow announced the succession of her favored heir, Servius Tullius.

These discoveries have divided classicists and archaeologists. There are some who view them as a brilliant, daring, and largely successful attempt to tie up the literary tradition with the physical remains, and as a triumphant demolition of all those with an “unjustifiably hypercritical attitude” who regard the myths of early Rome as just that, myths. Many others—and I am one of them—see instead a dangerous literal-mindedness in Carandini’s discoveries and reconstructions, somewhat akin to reported findings of King Arthur’s Round Table. The myths of Rome’s earliest phases (and there were many contradictory ones) are extremely important in our understanding of the Romans’ view of themselves, but Carandini provides no plausible account of how writers of the first century BC could access any reliable information about Rome of the eighth and seventh centuries BC, when there were no literary records of any sort (he relies entirely on the idea of “oral tradition”).

Besides, the closer you look at the particular claims about the individual sites, the flimsier they appear. The supposed “House of Tarquinius Priscus,” for example, hardly survives the scrutiny of T.P. Wiseman, who points out that there is very little remaining of this “house” at all (at most a few feet of possible wall), at the same time arguing convincingly, to me at least, that Carandini’s method of aligning the literary sources with the (largely nonexistent) remains is more wishful thinking than brilliant and daring.

Of course, the authors of the Atlas have every right to follow Carandini’s reconstructions if they choose, or to opt for whatever version of ancient Rome they want. But they have, I think, some obligation to acknowledge the disagreements and other views, which they very rarely do; there is, for example, no mention of Wiseman’s critique of the “House of Tarquinius Priscus,” although it appeared several years before the original Italian edition. It may well be that every reference book is driven by an ideology that it tries, usually unsuccessfully, to conceal (see the examples of word usage cited in the Oxford English Dictionary, or the Latin quotations used to illustrate grammatical rules in Kennedy’s Latin Primer); it is in fact almost the definition of a reference book that it is one that tries to conceal its ideology. But the Atlas is a particularly flagrant example of a partisan account masquerading as a work of reference, and given authority and credence by that. Again, reader beware.