Abstract

Studies of the sociology of contemporary earthquakes have emphasized the potentialities created by these disasters: earthquake-induced destruction, while traumatic, can also clear the way for large-scale infrastructural and architectural development programs with the potential to reshape aged urban environments and better reflect changing societal values and priorities. This chapter offers a survey of earthquakes as non-human change agents in the Roman and Late Antique Mediterranean, with special focus on the cities of Pompeii, Ephesus, Antioch, and Phrygian Hierapolis. While contemporary Roman sources tend to describe urban rebuilding after earthquakes in a symbolic manner, with a generic picture of cities “rebuilt” or “restored” and state-directed support sent for finance or labor, these literary images rarely correspond with the archaeological evidence for earthquake events in Roman cities, whose records leave little that speaks to the immediate challenges of search and rescue or mortalities but which also provided opportunities for the implementation of altogether new urban schemes.

The geological forces that create earthquakes are so colossal in scale and time that they are nearly incomprehensible from the perspective of our own short human lives. Yet the extraordinary violence of earthquakes transpires in mere seconds, with effects that permanently alter societies and communities. Earthquakes have always been and remain unpredictable, despite pre-modern efforts to understand their origins. Seneca the Younger’s Natural Questions (ca. CE 65) offered cautiously that we might “face this disaster bravely [because] it can neither be avoided nor foreseen,” before critically considering various hypotheses for the natural causes of earthquakes offered by Greek philosophers in earlier centuries—including the movement of water, fire, earth, and air (Seneca 2010:88). Approximately seventy years later, in Han Dynasty China, Zhang Heng invented a device for detecting and measuring earthquakes that remained in use there for centuries and may have reappeared or been reinvented independently in thirteenth-century Persia (Needham 1959:624–635). In Byzantine Constantinople, Anthemios of Tralles—one of the architects of Justinian’s daringly domed Hagia Sophia, completed in 537 CE but which collapsed the first time during an earthquake as soon as 558—is credited with an experiment that produced an artificial earthquake using steam power, a product of ancient interest in earthquakes’ destructive potential (Agathias 1975:141).

Earthquakes are endemic to the Mediterranean—whose waters sit astride the conjuncture of Asia, Europe, and Africa—even if the theory of plate tectonics (Mallet 1862; Wegener 1928) that explains them was understood only as recently as the 1960s (Blacket et al. 1965; Le Pichon 1968; Wilson 1963, 1965). The reality that human societies have, perhaps counterintuitively, crowded settlements along continental fault lines may be related to seismicity’s less appreciated and more positive consequences: namely, the elevation and depression of landmasses into mountains, islands, and basins that correlate well with biodiversity and the distribution of precious metals and building stone. The resultant hydrology encompasses both the relation of coastal montane uplift with oceanic precipitation patterns and the locations of drainage basins that provide water from rivers and lakes (Broodbank 2013:65–71). These factors together are particularly critical in the Mediterranean, where the tectonic conjunction of three continents has created conditions that both facilitated the development of human society and put it at perennial risk of seismic catastrophe (Herman et al. 2015).

Historical earthquakes can help us understand the agency of the environment in a broader global and historical framework, especially within the operational contexts of resilience and robustness or fragility, adaptation, and sustainability (Walker et al. 2004). While the scaled physical impacts of earthquakes have been quantitatively measured using the Mercalli intensity or Richter magnitude systems ever since those scales were devised in the 1930s, the localized and qualitative effects of a given earthquake depend very much on physical, social, cultural, and historical contingencies. Forms of settlement organization and architectural tradition shape background risk to seismic events. Communal cohesion and organization can affect the scale and efficacy of search and rescue or relief efforts in the immediate aftermath of an earthquake; there are also highly variable capacities of administration and private citizens who can direct funds, materiel, and labor to coordinate reconstruction and mitigation for future events in the longer term (Mordechai and Pickett 2018). Depending on the contours of these factors in any given historical or cultural context, the societal impacts of two geologically similar earthquakes—for instance, at 8.0 magnitude (along the upper range of recorded occurrences)—may vary considerably in different locations and sociocultural milieus (Özerdem and Jacoby 2006).

There are more than 200 documented occurrences of earthquakes in the ancient Roman archaeological and textual record, providing a rich opportunity to address human-nature dynamics in an ancient complex society. A fuller account of seismicity’s agency in Roman society—as opposed to notions of the environment as mere passive background, in the typically Braudelian sense (Moore 2003:432)—might begin with discussion of primary sources and archaeological evidence for the infamous earthquake that struck Italy in 62 CE, devastating Naples Bay and, with it, Pompeii and Herculaneum (figure 4.1).

We have two literary accounts of the earthquake that affected the region around Naples Bay and Pompeii in 62 CE, from which Pompeii was still recovering when Vesuvius erupted just a short time later, in 79 CE. Seneca (2010:87) wrote to his friend Lucilius, in a letter later compiled into the Natural Questions:

We have heard, Lucilius, excellent man, that Pompeii, the busy Campanian city, has been ruined by an earthquake, and all the neighboring areas have been badly affected. The coasts of Surrentum and Stabiae on the one side, and of Herculaneum on the other, meet at the city, encircling the open sea, which there retreats inland in a charming bay. This event happened during winter too, a period that our ancestors used to promise us was free from such dangers. Campania had always been nervous of this threat but had remained unharmed and had many times got over its fears; but this earthquake, occurring on February 5 [in the consulship of Regulus and Verginius], devastated all of the region and caused great destruction. For part of the town of Herculaneum collapsed too, and even what remains is standing precariously. The colony of Nuceria, though spared catastrophe, is not spared complaints. Naples, too, has lost many private buildings, though no civic ones, being only lightly affected by the vast tragedy. Country houses have collapsed or have often been shaken without damage. In addition to this, a flock of hundreds of sheep was killed, and statues were split apart; afterward some people wandered around in a state of shock and deranged.

Seneca’s account of the 62 CE earthquake at Pompeii is the longest on record for this particular earthquake, though it is only the introduction to one book of the Natural Questions’ much longer and decidedly modern consideration of the geological phenomena of earthquakes. Seneca here acknowledges the frequency of seismic activity around Naples—we know now that Naples stands on a major fault line—and he comments on the highly inconsistent, differential damage of houses ruined when public buildings were left standing. Tacitus (1931–1937:251), our only alternative literary source for this event, reports what happened there with the brevity typical of an annalist: “Pompeii, a famous town in Campania, was seriously damaged in an earthquake.” Through the reports of Seneca and Tacitus, we note several salient features of Greek and Latin seismicity, although accounts of Roman earthquakes often lack details of primary seismic effects (Seneca’s Natural Questions is actually an important exception, with his interest in phenomena such as surface faulting). Among other writers, details of catastrophic damage are sometimes only copied clichés—reports abound of whole cities destroyed or “swallowed up,” for instance—yet they are nevertheless adduced by modern catalogers and scientific observers to hypothesize earthquake magnitude or epicenter, often in the absence of corroborating archaeological or scientific evidence (Guidoboni 1994). In the context of Seneca’s account, for instance, much ink has been spilled in pursuit of a firm date for the Pompeii earthquake in 62 or 63 CE, though the former is now widely accepted.

Related is the matter of that earthquake’s epicenter and its magnitude: Seneca’s comparatively detailed account of structural damage following this earthquake, in combination with archaeological evidence (discussed below), have promoted estimates of a 5–6 Richter magnitude, considered a “moderate” seismic event in today’s terms. For the simple reason that Seneca’s description is quite explicit that Pompeii’s buildings suffered more damage than did those at Naples or Nucera, the epicenter is usually placed somewhere in the vicinity of Pompeii near Mount Vesuvius (Cubellis and Marturano 2013; Marturano 2008). Seneca and Tacitus are also (pace Agathias 1975:47, 137) little concerned with the secondary and tertiary or knock-on effects of the Pompeii earthquake, such as the organization of search and rescue efforts, or with the conflagrations, epidemics, and social unrest that flared alongside sharp declines in living standards in the immediate aftermath of the disaster.

Such highly literary reports of earthquakes should be understood in the context of Roman interests in environmental phenomena and anomalies, as well as in the frequently encountered suspicion that earthquakes were portents of sociopolitical crisis or manifestations of cosmic drama and divine wrath (Waldherr 2016; Deeg 2016). For example, the Ab Urbe Condita Libri (Livy ca. 9 BCE) and the anonymous Scriptores Historiae Augustae (ca. CE 350) are replete with this style of earthquake reportage for periods of uncertainty and heightened sensitivity to catastrophe, not least during the troubles of the Roman Republic or the crises of the empire during the third century CE, respectively. During the latter period, the Scriptores (1932:27) tell us that “amid so many calamities of war, there was also a terrible earthquake and darkness for many days . . . The disaster was worst in the cities of Asia; but Rome, too, was shaken, and Libya as well. In many places the earth gaped open . . . and many cities were also engulfed by the sea.”

The inspecificity of so many Roman earthquake reports, alongside the clichés of longer accounts, has too frequently meant that broader investigations of the social consequences of Mediterranean earthquakes have been neglected. Such factors together are critical for contemporary understandings of the nature of preexisting societal vulnerabilities to earthquakes and the recovery and rebuilding response that followed in ways we might define as resilient, adaptive, or transformative.

Resilience refers to specific ecological and social systems paradigms, which are commonly illustrated with the rolling ball-in-sink diagram (figure 4.2). Resilience here means “the capacity of a system to absorb disturbance and reorganize [or recover] while undergoing change so as to still retain essentially the same function, structure, identity and feedbacks” (Walker et al. 2004:5). Adaptation refers to the capacity of actors to manage resilience, providing a more robust return to the status quo ante but not changing the system itself. In contrast, transformation refers to the generation of an entirely new status quo with different ecological, cultural, or socioeconomic parameters following an event whose magnitude, in combination with existing vulnerabilities, exceeds a society’s ability to return to the status quo ante. As noted in the context of earthquakes, these differences are important because earthquakes that are objectively the same on the Richter scale have different effects depending on a society’s existing vulnerability or precarity, its preparedness and social cohesion, and its flexibility for either adaptation or transformation: we might compare the similar magnitudes but very different sociocultural effects of modern-era earthquakes in Tokyo and Istanbul to make this point (Özerdem and Jacoby 2006).

At Pompeii, inscriptions point toward the city’s resilience—the ability of citizenry and the state to organize and restore the city to its status quo ante—following the earthquake of 62 CE. An inscription from the Temple of Isis at Pompeii records its complete rebuilding with the resources of a local freedman: “Numerius Popidius . . . restored this building that collapsed in the earthquake from its foundations” (Corpus Inscriptionum Latinarum 1893). Much of the rebuilding work at Pompeii seems to have been organized locally, because decisive imperial or state-level intervention in the aftermath of the 62 CE earthquake at Pompeii is largely lacking in the historical record. There are no indications of immediate action or systemic responses taken here by the Roman emperor Nero, unlike the situation after the Great Fire of Rome in 64 CE when order was imposed on the city’s streets, with restrictions on height and requirements for stone rather than timber construction, modifications to water supply for fire protection, and so on (Tacitus 1931–1937 15.43). Imperial interventions at Pompeii after the 62 CE earthquake were limited instead to the single-building restorations attested by inscriptions found at the nearby city of Herculaneum, which date rather later, during the reign of Vespasian (r. 69–79 CE) (Corpus Inscriptionum Latinarum 1893; Année Epigraphique 1979:170). Not unrelated, a Vespasianic series of boundary-stone inscriptions called cippi may also attest to the intervention of imperially directed bureaucrats to confirm the limits of earthquake-damaged properties around the same time (Corpus Inscriptionum Latinarum 1893).

Pompeii’s archaeology provides an additional accounting of the city’s adaptability, with evidence for widespread rebuilding in the critical period between 62 CE and the city’s final destruction during the volcanic eruption of 79 CE (Adam 1989; Foss and Dobbins 2007). Buildings in Pompeii were to some extent, however incidentally, already built adaptively for earthquakes in symmetrical box-like fashion with limited height. Yet Pompeii also had substantial vulnerabilities. Buildings at Pompeii were typically constructed of very mixed or heterogeneous masonry, with multiple formats and sizes of stone or brick deployed within the walls of the same structure. The resultant variety of materials, each with its own natural frequencies of resonance, compounded risks for collapse during a seismic event (Adam 2005:290–318). Party walls between neighboring buildings were typically unbonded, while tie beams were used to undergird upper-story floors that were not secured across properties. These three common features of local building practice—inconsistent building materials, unbonded party walls, and disconnected tie beams—made Pompeian houses particularly vulnerable to toppling from rocking, horizontal seismic motions (Ruggieri 2017a, 2017b; Ruggieri et al. 2018). This point is acknowledged by the Roman author Seneca (2010:87) when he describes how private housing stock suffered more damage than did public buildings. In this way, local building practice had tremendous repercussions for lower- or middle-income residents living in rented-out rooms on upper floors, which were susceptible to collapse when exterior walls toppled (Scobie 1986; Ruggieri 2017a, 2017b). At the same time, column drums in temples and peristyles were also subject to toppling if they were not firmly fixed to the surface on which they stood. All of these risk factors were compounded by urban density and Pompeii’s proximity to the tectonic fault that runs north-south through the Italian peninsula.

Structural repairs and modifications in the years after the 62 CE Pompeii earthquake seem to have addressed some of the city’s preexisting vulnerabilities. Widespread cracking and separation in weak walls were repaired by replacement with either similar masonry or new brick-faced concrete called opus testaceum, especially in corners and piers. Opus testaceum was more resistant to horizontal motions with greater friction than were solid stone ashlars or mixed-format brick-and-stone masonry between courses in addition to being speedier to install thanks to standardized, mass-produced materials (Ruggieri 2017a, 2017b). Column drums across the town were fitted with bronze or lead pins that bonded them to sidewalk surfaces (as at Palestra Grande, Adam 2005:310). And perhaps resulting from a similar criterion—namely, improved resistance to horizontal seismic motions—domestic architecture at Pompeii was also retrofitted with masonry walls installed at around half the height of the columns (as at Villa of the Mysteries) or with buttresses on exterior walls (as at Villa di Diomedi) (Ruggieri 2017a, 2017b).

People at Pompeii were perceptive and innovative in their adaptations to the inherent risk presented by settlement in the seismically active areas along the fault that runs the length of the Italian peninsula. Such adaptations are not unique to Pompeii but in fact appear globally—at different times, in different places—as common architectural responses to seismic risk (Ortega et al. 2017). At the same time, we should not overlook the fact that such adaptations—which perpetuated human settlement in an endangered place and promoted resilience and restoration with clever changes in building practice—were not transformative for institutions or ways of living; nor were they ultimately sustainable for the city’s inhabitants. Pompeii was, of course, buried in ash and lava after the 79 CE earthquake triggered the eruption of Vesuvius and brought the city’s existence to an end (Sigurdsson 2007).

Indeed, today’s environmental historians, urban planners, and students of historical disasters have emphasized the differences between adaptation (incremental innovations that accommodate increased stresses to maintain system functionality) and transformation (disruptive innovations implying long-term transitions) to note, for instance, that “the current political arena favors adaptation because it works to maintain the established order and address near-term problems” (Redman 2014:38) while framing neo-liberal urbanism as one “example of how highly . . . unsustainable and inequitable systems can be extremely resilient to change” (Chelleri 2016). Pompeii would never have been sustainable in the long run by virtue of its location—a city built on a fault under an active volcano—but it was arguably resilient to 5–6 magnitude earthquakes. On the other hand, Romans had no historical memory of Vesuvius’s last eruption 700 years before 79 CE (Sigurdsson 2007:45). Following the latter catastrophe, the city was never rebuilt, even if the region around it flourished in the centuries to come. Indeed, Penelope M. Allison (2002) has argued that the absolute impacts of the loss of Pompeii were surprisingly negligible in terms of the contemporary Roman economy. We might ask, as the Romans may have, too: were the short-term, adaptive gains of rebuilding after the 62 CE earthquake worth the long-term loss following a volcanic eruption? Was Pompeii’s response to earthquake typical of other Roman cities faced with seismic threats? Was abandonment the only truly transformative, sustainable approach to settlement in Pompeii? In other Roman cities built alongside a fault but without the threat of an active volcano, how might we recognize resilience, adaptation, or transformation?

Any assessment of the nature of Roman response to earthquakes—whether consistently resilient and adaptive with returns to the status quo ante or indeed transformative and perhaps even sustainable—must depend on a larger dataset for seismic events across the Roman Mediterranean, including scientific, archaeological, and textual information. Among these, the corpus of Roman and Late Antique literature and historical attestations of seismic events is still our fundamental and most comprehensive resource. Even in large and long-excavated Roman cities, where earthquakes known from literary sources have long been evoked as agents in narratives of historical change, earthquake damage is only beginning to be recognized and studied scientifically (Rodríguez-Pascua et al. 2011).

A projection of 265 seismic events (figure 4.3) recorded by Roman authors for the years between 200 BCE and 800 CE—compiled from the catalogs produced by Emanuela Guidoboni (1994) and Nicholas Ambraseys (2009)—enables us to visualize several important patterns. It is clear, for instance, that earthquakes reported by ancient literary sources do indeed follow the major Mediterranean fault lines (as indicated in figure 4.3 by solid lines) and that reported earthquakes are clustered in zones of population density. Note, for instance, how earthquakes are densely spread along the Dead Sea transform fault, which connects the Red Sea to the Taurus Mountains, and along the fault as it bends back west through Cyprus—another hotspot—before splitting north along the heavily urbanized and seismically active western Anatolian coast, as well as west toward Crete and Greece and north through Italy into the Alps.

Population size and political importance appear to be the primary factors in the distribution of seismic events attested among Roman primary sources. Provincial cities in the Levant, the coastal Aegean, Crete or Cyprus, and southern Anatolia constituted the primary tax base for the Roman empire, especially between the third and seventh centuries CE. Elsewhere, locales of reported earthquakes reflect the interest of court cultures or particular authors much more than they represent any absolute record of seismic events as they actually occurred, as would be the case for modern instrumental records. For instance, Greece and its environs pick up their seismic signal largely as the product of the second-century CE geographer Strabo, whose attention to anomalies of nature in places of mythological significance was second to none. Later, earthquakes were well reported for the coastal Italian city of Ravenna in the fifth century CE—about which there was no mention before this period—during its brief fluorescence as capital of the Ostrogothic kingdom (and, consequently, the center of activity for royal chroniclers there).

Although accounts of individual Roman earthquakes are markedly uneven, taken altogether, they cohere as a consistent set of Roman expectations and responses congruent with the contours of disaster and disaster response as described by contemporary scholarship: primary environmental and geological effects on the landscape and architecture when a seismic event occurs; secondary-stage social effects immediately thereafter, with search and rescue alongside fire, food shortages, and other dramatic changes in living standards and habits; followed eventually by tertiary rebuilding and reconstruction or mitigation efforts that could, to varying degrees, reshape entire urban landscapes and communities or ways of living.

The immediate responses to earthquakes emanated primarily from local capacities. As a result of slow lines of communication, funds sent by emperors from the capital in the initial aftermath of an earthquake inevitably came too late and could serve only to aid in the recovery of the dead. Witness Anastasius’s response to Rhodes in 515: “The merciful emperor sent much gold to those who were found still alive on the island. They started to dig out and extract the bodies of those who had been squashed, as if in a wine-press, by this great cataclysm” (Pseudo-Dionysus 1996:6). Just so, Dio Cassius reports on ill-fated search and rescue efforts made after an earthquake and its aftershocks struck Syrian Antioch in December 115 CE, when the emperor Trajan happened to be passing through the city en route to the Parthian front. Even though Trajan was in a position to immediately bring to bear the resources of the Roman army and state upon the devastated city, these efforts seem to have been unsuccessful:

Heaven continued the earthquake for several days and nights, the people were in dire straits and helpless, some of them crushed and perishing under the weight of the buildings pressing upon them, and others dying of hunger, whenever it so chanced that they were left alive either in a clear space, the timbers being so inclined as to leave such a space, or in a vaulted colonnade. They searched the . . . heaps [of ruined buildings], but were not able to find in them anyone still living save a child sucking at the breast of its mother, who was dead. As they drew forth the corpses they could no longer feel any pleasure even at their own escape. (Dio Cassius 1925:405)

Such accounts constitute the bulk of our information about Roman search and rescue or charitable operations in the aftermath of disasters, which depended entirely on local-level ad hoc arrangements that are often, if only schematically, described by our sources.

To add another unfortunately detailed example, Ammianus Marcellinus (1935–1940:341–345) reports the primary geological and architectural effects of an intense earthquake at Nicomedia in 358 CE, with remarkable (if occasionally clichéd) details, before vividly describing search and rescue efforts—“the highest points [of the city] re-echoed all manner of outcries, of those seeking their wives, their children, and whatever near kinsfolk belonged to them”—that were catastrophically disrupted by a conflagration that burned the beleaguered city’s remains. For modern disaster sociologists, community cohesion is key to survival in the hours and days after a seismic event. The vast majority of trapped victims brought out alive from beneath the rubble are rescued in the first twenty-four hours after an event, usually by neighbors and family rather than specialists (Smith 2013:117). Survivors’ participation in community-sponsored offerings of thanksgiving, including even the construction of entire temples, may have fostered such community cohesion in future earthquakes: following an earthquake at Antioch in 115 CE, “the surviving Antiochenes who remained then built a temple on which they inscribed ‘Those who were saved erected this to Zeus the Savior’ ” (Malalas 1986:145). Early Christian histories record comparable community events that commemorated earthquakes, incorporated into the annual calendars of the church, which were focused on public liturgies with prayers and processions through frequently afflicted cities like Constantinople—led by authorities that could include the emperor, the patriarch, and his bishops (Croke 1981; see similar activities associated with plague outbreaks discussed in Horden 2008).

Compared to recovery and rescue operations, we are much better informed about the nature of Roman reconstruction activities, at least insofar as such efforts were directed from the capital and recorded by historians. Despite the agency and importance of local communities, rulers who embodied the state or their agents received much of the credit for post-earthquake reconstruction. Textual sources reveal a considerable range of administrative options that evolved to finance reconstruction activities and which testify to the adaptability of the Roman state and its entire bureaucracy under stress. Chief among these options, however, are direct contributions from the state fisc at an emperor’s direction or with a senatorial resolution (Dio Cassius 1927:57), as well as imperially financed reconstructions supervised by local officials (e.g., Année Epigraphique 1913:227). The provision of construction materials or skilled labor (including engineers and soldiers) was probably of comparable significance with funding for earthquake-damaged communities, if also less easily quantifiable and typically absent or mentioned in passing by Roman sources. Local labor is never mentioned.

Although Roman authors emphasized the importance of direct state contributions for earthquake-damaged cities, imperially allocated funds could be quite limited in comparison with costs for new construction projects during the same period. For instance: among the primary sources describing monetary relief in the sixth century, Malalas (1986:419–422) indicates that 500 pounds of gold were sent to Antioch for reconstruction after the 526 CE earthquake, and 200 pounds of gold were sent to Laodicea after an earthquake in 528 CE (443). Under normal circumstances, such sums were considerable, but they could also be obtained by a single city through its own resources or even by a single individual. During the earlier sixth century, for example, we read that a squadron of merchant ships operated by the church of Alexandria sank in the Adriatic with cargoes worth 3,400 pounds of gold, that the title of Archbishop of Alexandria was purchased for 700 pounds of gold (Monks 1953), and that just one building—the resplendent new Church of San Vitale in Ravenna, famous for its mosaics and sculpture—was paid for by a banker and valued at approximately 400 pounds of gold (Agnellus 2004:59).

Another related option by which the Roman state might finance reconstruction after an earthquake was forgiveness of a city’s entire tax bill for a temporary period, often five to ten years. Such tax remittances following earthquakes could be substantial, perhaps more so than direct contributions of coin. For instance, the emperor Tiberius granted that 10 million sesterces were to be wiped clean from the tax bill owed by the city of Sardis in 17 CE (Tacitus 1931–1937:25). In later centuries, after an earthquake at Antioch in 458 CE, we are told that approximately 100,000 pounds of gold were kept from the hands of imperial tax collectors (Evagrius Scholasticus 2000:96). Property owner taxes on individual buildings lost to earthquake damages were also occasionally remitted, or particular forms of tax or service could be suspended for varying lengths of time (e.g., Claudius’s suspension of a hearth tax at Antioch, with money redirected to reconstruction of street colonnades and porticoes; Malalas 1986:131). Some communities were also reportedly capable of rebuilding from their own funds, without imperial intervention (Tacitus 1937:151), while others organized donations and incentivized reconstruction of individual structures by local notables (Corpus Inscriptionum Latinarum 1893). For example, Opramoas of Rhodiapolis gave huge sums to thirty cities in southern Anatolia after earthquakes there in 141 CE, as recorded by a famous inscription (Petersen and von Luschan 1889:2, 109–115).

Having reviewed some of the finances of Roman reconstruction after earthquakes, we might inquire into the qualitative nature of rebuilding and restoration following seismic events—which, to judge from surviving sources, tended to focus repair efforts on ideologically important buildings such as temples and churches rather than infrastructure or housing (Mordechai and Pickett 2018). Roman sources described post-earthquake urban reconstruction in a decidedly symbolic manner, with a generic picture of entire cities “rebuilt,” “reconstructed,” or “restored.” However, these rhetorical images—wherein cities were recorded as having been returned to their pristine pre-disaster state—rarely correspond to the archaeological evidence.

Three Case Studies of Post-Earthquake Reconstruction

A comparison of how seismic events reshaped cities and ways of living in three Roman cities—Syrian Antioch, Asian Ephesus, and Phrygian Hierapolis—highlights the intriguing disjunction between “restoration” or resilient returns to the status quo ante that are so prominent in literary sources and the facts of more transformative rebuilding episodes as they are known archaeologically (Mordechai and Pickett 2018:341–343). Post-earthquake reconstruction episodes known from archaeology indicate that whatever the literary sources might tell us, Romans rarely “rebuilt” cities in a literal sense; rather—like Lisbon after the 1755 earthquake (Mullin 1992; Chester 2001; Ribeiro dos Santos 2011)—earthquakes created spaces of institutional and architectural potentiality that were not merely adaptive and resilient but which rose to the level of transformation through a process of provoking or enabling reorganization of institutions and communities from the ruins of cities.

Discussion

Despite images of basic infrastructural resilience at Hierapolis and Ephesus during the fourth and fifth centuries—accompanied by substantial transformations in the cities’ fabrics that reflected deep-seated societal evolutions over the previous decades—we are hard-pressed to define these changes as resilience or transformation unless we also ask, change for whom? Earthquake-induced changes at Hierapolis and Ephesus did not hinder the state-level functionality of either city as a center for tax collection, always the primary desideratum of the state, through regulation of trade or hinterland resources. The frequency and chronology of coin finds in those two cities could suggest a reflorescence of economic activity in the years following earthquakes. But should we view renewed activity as adaptations? Earthquakes transformed late Roman cities in ways that may have been unsettling to urban elites, obliterating traditional venues and forms of public display and munificence (e.g., portraits and inscriptions at temples or the theater) or repurposing them to more mundane, if economically satisfying, functions of lower- and middle-class artisanal housing. From these perspectives, earthquake-induced reorganizations of urban form and function might well be labeled as transformative for members of all social strata.

Ambiguities of resilience, adaptation, and transformation resolved with a cluster of devastating earthquakes in the late seventh century that pushed Hierapolis into a different state entirely (figure 4.4). This transformation—which excavators there strongly associated with a seismic event ca. 670 CE that is well evidenced archaeologically but entirely absent among Roman literary sources—included the total abandonment of the city’s public architecture, with its water supply systems and rectilinear streets. Low-density residences and industrial installations moved into the shells of collapsed buildings in the area around the city’s unrepaired walls, while the St. Philip complex outside the walls was repaired and seemingly continued to function as a necropolis and a focus of regional and supra-regional pilgrimage long after the city’s other churches had collapsed. These seismically triggered changes constituted a radical, low-cost form of reorganization of settlement patterns and re-focus within the remains of the ruined antique city, which survived in this form for six centuries until the arrival of the Seljuks in the Lycos valley around Hierapolis in the eleventh century CE.

Conclusion

Seismically driven changes were endemic to the urban landscapes of the eastern Mediterranean during the Roman and Late Antiquity eras, symptoms of deep structural changes in Roman society made manifest in spaces of potentiality. The role of earthquakes in urban change during this period should be understood as only one pendulous contributing factor, whose magnitude could punctuate or accelerate longer-cycle evolutions and thereby push urban systems, such as Antioch or Ephesus or Hierapolis, into a transformed condition. Seismic events in the Roman world were not merely occasions for rebuilding and resilience, with adaptations that could return urban societies to the status quo ante; earthquakes also provoked, enabled, and formalized longer-term urban and social transformations.

References