Late Roman Economy and Formation Processes

I’ve spent some quality time with the most recent volume of Late Antique Archaeology this past month in preparation for writing a short contribution with David Pettegrew on connectivity in the Late Roman eastern Mediterranean. We plan to compare the Late Roman assemblages produced by two survey projects:  Eastern Korinthia Archaeological Project and Pyla-Koutsopetria Archaeology Project. An important component of both assemblages is Late Roman amphoras: EKAS produced substantial quantities of Late Roman 2 Amphora probably produced in the Argolid; PKAP produced quantities of Late Roman 1 Amphora produced both on Cyprus and in southern Cilicia. We hope to discuss how the concentrations of these common transport vessels reflected and complicated how we understand economic patterns in the Late Antiquity.

Over the past half-century two basic models for the Late Roman economy have emerged. The earlier models saw the state as the primary engine for trade in antiquity. More recently, however, scholars have argued that the core feature of ancient trade is small-scale interaction between microregions across the Mediterranean basin. While there is undoubtedly some truth in both models, the latter has substantially more favor among scholars at present and the volume dedicated to connectivity focuses on the kind of small-scale interregional exchange that created a network of social, economic, and even cultural connections that defined the ancient Mediterranean world. The classic question introduced to complicate our view of ancient connectivity is: if the ancient Mediterranean is defined by these small-scale connections, then why did the political, economic, social, and even cultural unity of the communities tied to the Middle Sea collapse with the fall of Roman political organization in Late Antiquity?

Figure4 18

This is where David and I want to introduce the complicating matter of formation process archaeology. The substantial assemblages of Late Roman amphora represent the accumulation of discard from two “nodes” within the Late Antique economic network. These two nodes, however, are particularly visible because of the substantial concentration of a class of transport vessel.

These transport vessels most likely served to transport supplies to imperial troops either stationed in the Balkans or around the Black Sea, or in the case of the Eastern Korinthia, working to refortify the massive Hexamilion Wall that ran the width of the Isthmus of Corinth or stationed in its eastern fortress near the sanctuary of Isthmia. The visibility of these two areas depends upon a kind of artifact associated with a kind of exchange. As David has noted the surface treatments associated with LR2 amphora make them highly diagnostic in the surface record. LR1s, in turn, have highly diagnostic, twisted, handles that make them stand out from a surface assemblage dominated by relatively undifferentiated body sherds. In other words, these amphora assemblages represent a visible kind of economic activity.

The impact of this visible type of economic activity on our understanding of Late Roman connectivity is complex. On the one hand, the kind of persistent, low-level, economic connections associated with most models of connectivity are unlikely to leave much evidence on the surface. The diverse and relatively small group of very diverse amphoras, for example, found upon the coasting vessel at Fig Tree Bay on Cyprus would have been deposited at numerous small harbors along its route. Moreover, the fluidity of the networks that characterized connectivity would have made the routes of caboteurs irregular and contingent on various economic situations throughout the network of relationships. This variability and the small-scale of this activity is unlikely to have created an archaeologically visible assemblage at any one point on these routes. More than this, overland trade in wine or olive oil may not have used amphoras at all further impairing the archaeological visibility of the kind of low-level connectivity characteristic of Mediterranean exchange patterns. Between ephemeral containers and variable, low-density scatters, the regular pattern of archaeological exchange characterizing connectivity will never be especially visible in the landscape.

In contrast, imperial provisioning requirements, fueled for example by the quaestura exercitus, would present exceptionally visible assemblages of material. The interesting thing, to me, is that the amphoras visible on the surface in the Korinthia and at Koutsopetria  are not what is being exchanged, but the containers in which exchange occurs. The material exchanged, olive oil and wine, are almost entirely invisible in the archaeological records on their own. The visibility of these two places reflects the presence of outlets for a region’s produce. The produce itself, however, leaves very little trace, and we have to assume that networks that integrated microregions across the Mediterranean functioned to bring goods from across a wide area to a particular site for large-scale export.

The collapse of these sites of large-scale export during the tumultuous 7th and 8th centuries did not make trade between microregions end, but it made it more contingent and less visible, as I have argued for this period on Cyprus. The absence of large accumulations of highly diagnostic artifact types in one place represent a return to our ability to recognize normal patterns of Mediterranean exchange as much as the disruption of this exchange. The decline of these sites both deprived archaeologists of visible monuments of exchange and ancient communities of a brief moment of economic stability within longstanding contingent networks.

Connectivity on Cyprus and Corinth

David Pettegrew and I are working up a paper for a volume on connectivity in the ancient Mediterranean. Connectivity has been a buzz word in Mediterranean archaeology since Horden and Purcell’s The Corrupting Sea used it to describe the regular pattern of small-scale connections between microregions. These microregions depend upon connectivity for political and social stability and economic subsistence.  

Our original plan was to compare the artifact assemblages at our two research sites on Cyprus: Polis-Chrysochous and Pyla-Koutsopetria and show how these two sites engaged the broader Mediterranean world in a different ways. They not only showed links to different regional networks of exchange, but also showed different kinds of relationships to these networks. Polis, for example, was a small city and Koutsopetria seems to have been a regional emporium directed toward the export of agricultural goods. 

After mulling this paper over for a few weeks (and missing some deadlines and conjuring enthusiasm for various arguments), we decided to take a shot at making a very generous deadline extension and turn the paper in a different direction. David is almost finished his book on the history of the Corinthian Isthmus based heavily on the work of the Eastern Korinthia Archaeological Survey (EKAS) and we have also recently submitted our completed manuscript documenting our intensive survey at Pyla-Koutsopetria. So it occurred to us that we might productively compare the results from these two survey projects as they share methods and sampling strategies.

More than that, the assemblages produced by comparable methods have certain clear similarities. Both study areas produced an abundance of Late Romam material particular easily-identified Late Roman amphoras. In the case of Koutsopetria, these are largely Late Roman type 1 amphora. In the Corinthia, the survey area produced a substantial quantity of Late Roman type 2 amphoras. While neither amphora was produced locally, both are regional types and LR1 kilns are known on Cyprus and there are LR2 kilns in the Southern Argolid. Both of these amphora types have been associated with forms of administrative trade in the Late Roman world, and provisioning the army on the borders of the empire in particular. 

Connectivity has tended to focus on the small-scale trade between interdependent microregions rather than the larger-scale, administrative trade. In fact, considering the role of this larger-scale trade in our notions of connectivity marks a return to older notions of trade in the Late Roman world which saw economic activity largely stimulated by the requirements of supplying the capital and the armies. The Corinthian Isthmus featured both imperially funded construction in the Hexamilion wall and, at least in the 6th century, a garrison of troops at fortress at Isthmia. The appearance of LR2 amphora in this context suggests the movement of goods into the area most likely to provision the garrison and to supply construction crews associated with the Hexamilion wall renovations in the 6th century.

At Koutsopetria, the abundance of LR1 is perhaps tied to the need to supply the army in the Balkans. The site may have served as a transshipment point for agricultural produce leaving Cyprus through the small embayment there. The numerous fragments of amphora there makes it unlikely that they represent goods coming into a small community, but more likely represented exports. The uniformity of the amphora types also suggests that goods are flowing out from the site in a systematic way.

The advantage of comparing these two study areas is to present a useful counterpoint  to the common view of connectivity that emphasizes links between microregions. Our paper will return to a view of the Mediterranean that considers the links between small places and the center while at the same time attempting to understand how these connections influence their relationships to other small places in their regions.

Archaeology of Sound

Every now and then when I’m in the field, I panic about falling behind in my journal reading and letting the ENTIRE DISCIPLINE PASS ME BY.

WHAT?? Archaeological Dialogues has an issue dedicated to ROMANIZATION? I thought about that once, like four years ago! I must… read… now!

WORLD ARCHAEOLOGY has forthcoming volume dedicated to the archaeology of sound? I know people working on that RIGHT NOW and how can I possibly interact with them without being familiar with soon-to-be-published articles. More than that, I’m an audiophile and I need to understand the archaeology of connectors. And I’ve done archaeology of the contemporary world (forthcoming) so I must understand what was albums were found on the floor of a commune where the Grateful Dead once live.  

It’s not that it has to happen eventually – like say while I’m on sabbatical – it has to happen now.

So instead of spending a weekend catching up on vital scholarship and remaining relevant to my discipline, I decided to clean up some audio file that I captured over the past few weeks in the field.

On my hike to the cave, I encounter a fairly agitated hawk and this what he (or she) sounded like:

We’ve also had the good fortune of encountering some very vocal goats:

And some excitable frogs (especially at night!):

Finally, you can faintly hear the bells of the church at Kaparelli at the western edge of our survey area:

A Bridge

This is mainly to start a blog post with the line that I want you use at the beginning of an important article:

“The study of Ottoman bridges in the Western Argolid remains in its infancy. The goal of this brief article is to bring attention to a small, but important body of Ottoman bridge work in this region.”



This lovely arch spanned a small ravine and carried a switchback kalderimi road down a low saddle to the village of Lyrkeia and our survey area. The stone work is lovely consisting of local grey limestone faces with smaller stones used as chinking. The arch itself is made of thinner stones arranged carefully with a substantial quantity of pebbly white mortar.

The road that leads to this bridge runs on its own carefully wrought terrace through olive groves. The is evidence that the bedrock had been cut back to let the road pass more easily. The bedrock was close enough to the surface to allow it serve as paving for part of the route, and it probably made this particular field appealing for use as a road (and less than appealing for agriculture!).

Late Antiquity in the Western Argolid

The Western Argolid Regional Project has the distinct benefit of two senior staff members who specialize in Late Antiquity. Scott Gallimore, one of the co-directors, recently finished a dissertation on Late Roman Crete, and people who read this blog should be pretty aware of my interest in that period

This concentration on the Late Antiquity is, at first blush, appropriate for a project in the Argolid which scholars have long understood to be a center of activity in this period. The city of Argos, for example, appeared on most of the prominent Late Antique geographies, and had a prominent bishop in Late Antiquity who attended the council of Constantinople in 381, Chacedon in 451, and Constantinople II in 680. The ancient city was riddled with Early Christian basilicas, cemeteries, and mosaic fragments of Late Roman date. So-called, “slavic” pottery, appeared in Argos suggesting that it saw a change in material culture consistent with sites elsewhere in the northeastern Peloponnesus.

Outside of Argos, there is evidence for rather intensive activities throughout the coastal region of the Argolic Gulf. The village of Myloi, where we stay, produced a Late Roman building, probably an exurban villa, of Late Roman date, and Late Roman activity extended inland from there around the village of Skaphadaki. Across the Gulf, Nauplion produced inscriptions of the 4th century (and an informal walk through town reveals spolia of Late Antique date) and a villa was discovered near the site of Asine – better known for its earlier remains. In the well-explored Southern Argolid, Halieis and Hermione witnessed signifiant activities in Late Antiquity with the former a production center for Late Roman 2 amphora and the latter featuring a Early Christian basilica complex with impressive mosaics and inscriptions mentioning a bishop Hermias. Troezene appears in Hierokles and was a center of ecclesiastical activity with a basilica and inscriptions, and despite its coastal location it appears to have survived into the 8th century with a bishop appearing at the Second Council of Nicaea reinforced with evidence from seals. The churches in the area of Epidauros are well-known and long thought to be among the earliest in Greece (on the dubious basis of architectural style). At Ano Epidauros a substantial quantity of Late Antique activity appeared, including the intriguing church at Lailoteika which may date to the 7th century or later. Scholars have long debated the reason for the Late Antique flourishing of activity on the small islands of the Saronic Gulf like Spetses, Dolkos, and Chinitsa which seems to have continued in the 7th century.   

The Late Romans did not spare the Argolid’s famous Bronze Age sites, with the neighborhood of Limnes, Prosymna, and the mighty Tiryns producing Early Christian graves and the citadel of Midea featuring activities in the 5th or 6th centuries. 

To use a vivid Appalachian saying: you can’t swing a dead cat without hitting Late Roman or Early Christian remains in the Argolid.

In contrast, the valleys of the Western Argolid including our survey area which follows the upper reaches of the Inachos River from the village of Kaparelli east through Lyrkeia and ancient Orneai, toward Sterna and the northwestern suburbs of  Argos. This region is a blank space without almost no published sites of Late Roman date. In fact, the most prominent Late Roman site in our survey area appears in a two-page reference to some Early Christian remains around the village of Lyrkeia by Dimitrios Pallas in the ADeltion of 1960 (pp. 100-101). 

Needless to say, this is odd. The valley bottom is fertile and the river provided a transportation route between the densely settled Argive plains and Arcadia which continued to prosper at least judging from the numerous buildings of Late Roman in this region. Moreover, the (relatively) easily traversed passes, strategic hill tops, and accessible valley walls, presented exactly the kind of topography to attract the attention of Late Roman military planners. This kind of “marginal land” also tended to attract Late Roman settlement. Recent scholarship has seen 5th and the first half of the 6th centuries as a period of population growth and settlement expansion manifest in monumental architecture and extensive trade in easily recognized ceramic types. In other words, the upper Inachos valley is exactly the kind of place where you’d expect Late Roman activity.

Settlement on Cyprus in the 7th and 8th centuries

This past week R. Scott Moore and I sent off a draft of a paper on settlement in Cyprus during the Early Byzantine period. 

This has been a work in progress for the last few months and developed partially from our work on Cyprus at the Pyla-Koutsopetria Archaeological Project and at the site of Polis-Chrysochous. 

This is the earlier draft of this paper, and you can read more on this paper hereherehere, and here.

The Northern Levant at the End of Antiquity

I was pretty excited to read Jesse Casana’s very recent article on the Late Roman landscape of the Northern Levant in the most recent Oxford Journal of Archaeology. I’ve been poking, in a tentative way, around this region lately (via articles and books, mind you) in an effort to situate Cyprus more clearly in its regional context. Casana’s article was particularly insightful because he relied heavily on evidence from survey archaeology.

He drew upon a number of recent survey projects (Ghab Regional Survey and the Amuq Valley Regional Project) to demonstrate that the lower Orontes Valley in the immediate neighborhood of Antioch was densely settled throughout the 5th and 6th centuries. The settlements on these fertile valley bottoms have largely been overlooked by scholars of the Late Roman period distracted, it would seem, by the dramatic remains of the “Dead Cities” of the limestone massif some 20 km to the west. The Dead Cities are remarkably well-preserved largely because the relatively arid landscape of the limestone massif was not reoccupied in later periods leaving the substantial limestone structures standing until today.  Moreover, scholars working in the Orontes Valley tended to study  the prominent tell sites which primarily date to the Bronze Age and Iron Age and overlooked the scattered tiles and ceramics that provide evidence for Roman and Late Roman occupation of these regions.

The evidence for Late Roman occupation in this region was substantial and, as Casana documented in a small-scale excavation, included elaborate buildings whose walls were either robbed for building material in later times or were made of mud brick. Casana argues that these apparently affluent settlements developed in response to markets in Antioch, Apamea, and accessed by sea from the coast of the North Levant. The Dead Cities, occupying more marginal land, are part of this same process of producing for booming urban markets and dynamic regional trade.  

Casana’s understanding of the boom in the Orontes Valley coincides with my reading of settlement on Cyprus. The Late Roman period in the East – perhaps into the 7th century on Cyprus – represented a period of urban prosperity, a high degree of monetization, and thriving regional markets in the Eastern Mediterranean stimulated at least, in part, through imperial policy and the needs of the army on the frontiers and the capital at Constantinople. The opportunities of the market stimulated the exploitation of marginal lands and this coincided with a gradual diversification of agricultural production from strictly subsistence practices to limited, opportunistic production for market. As Michael Decker has argued for the same region marginal lands sometimes become opportunities for niche production and the traditional reading of the Dead Cities on the limestone massif suggested that these villages produced olive oil primarily for export (although more recent work has shown that the villages may have also produced wine and grain perhaps for local consumption). 

As a conclusion, Casana frames the issues involving the structure of settlement in the Northern Levant as primarily archaeological in character. In other words, the remarkable preservation of the Dead Cities of the limestone massif has led scholars to overlook and mischaracterize contemporary settlement on the more fertile lands of the Orontes valley. This, as one can imagine, distorted the reading of settlement in this region and overlooked the massive expansion of settlement present in the region. The work of the two surveys summarized by Casana brings the Northern Levant in line with contemporary settlement patterns in the so-called “busy countryside” of Late Roman Cyprus. Like the Northern Levant, the booming urbanism of Late Roman Cyprus and access to the substantial and monetized Eastern Mediterranean economic world supported the expansion of settlement across the island. When the cohesive Eastern Mediterranean market faltered in the face of invasions and plagues in the later 7th century (on Cyprus and perhaps in the Levant as well), urban areas declined and regional markets returned to levels prior to the momentary stimulus provided by the state and an exception period of economic and political integration.     

A Working Paper on Settlement on Cyprus in the 7th and 8th Centuries

Since we’re in excavation mode today after our spring storm yesterday, I’ll offer up a working paper for your inspection. The paper looks at settlement on Cyprus during the 7th and 8th centuries and argues three interrelated things.

First, it challenges the idea that scholars should consider Late Roman Cyprus to be normative and late-7th and 8th century Cyprus in decline. Cyprus during 5th-early 7th century experienced a rather extraordinary period of economic prosperity and economic integration with both the Roman state and markets across the Mediterranean basin. 

Next, I suggest that our inability to grasp the situation on Cyprus during the late 7th and 8th century (as well as our tendency to declare Cyprus in decline) exists at the intersection of archaeological methods and historical circumstances. Our tendency to rely on artifact that circulated widely in the Mediterranean to establish chronological control of sites on Cyprus and our inability to consistently recognize locally produced ceramics (except those produced Cyprus that circulated widely), created a situation where disruptions to the Mediterranean economy in the late 7th and 8th centuries disrupted our ability as archaeologists to study communities on Cyprus.

Finally, I follow the well-trod path of a gaggle of recent scholars who have suggested that there is evidence for continued economic activity in the late 7th and 8th centuries across the island, but we need to become more sensitized to the kinds of evidence present. Arab coins, handmade pottery, and late forms of well-known, earlier types will tell the story of a dynamic, persistent, if more contingent economy, during the so-called “Dark Ages.”

More on this paper here, here, here, and here.   

Or you can just read the working paper here:

Writing as Process and the 7th Century on Cyprus

Spring break is one of my favorite times in my busy semester because for the last few years, I’ve been able to dedicate this time to a sustained writing project. In a normal semester, my writing has tended to get broken into tiny fragments of time – a morning here or an afternoon there – between teaching responsibilities, service, and other faculty duties. The result of this situation is that anything I write tends to be highly granular and composed of tiny 300 word snippets cobbled together and smoothed over in editing.

This is fine in some circumstances, but is hardly conducive to producing sustained and careful arguments. Spring break writing (and this goes for winter break and late, post-field season, summer writing too) holds forth the elusive opportunity to write in a series of 1000 word chunks over the course of five consecutive days (until a wife-mandated “rest day/date night”!). 

What this sustained writing has helped me to see are the little strands that make intuitive connections communicable. For my paper on Cyprus in the 7th century, for example, I’ve been able to notice the arguments for the appearance of handmade pottery in 7th century contexts on Cyprus and the disappearance of large issue coins are  interrelated in my argument. Handmade pottery appears in assemblages alongside both imported fine wares and locally produced cooking wares indicating that it was not a response to the abrupt end of regional or local trade and production. Instead, it would appear that handmade table and utilities wares appeared on Cyprus in a gradual way as a local response to slowly changing pattern of access. As for coins, the disappearance of large issues on Cyprus has sometimes been seen as evidence for abrupt economic decline, and there is little doubt that the disappearance of large issues after the reign of Constans II indicates some kind of economic change on the island. At the same time, Guy Sanders has noted that we are likely missing many of the small issues (nummi or minimi) that circulated throughout Late Antiquity because they were so small that they slipped through the excavator’s sieves. Like handmade pottery, these tiny coins served to shape Late Antique life on Cyprus in a way not entirely visible to the 20th century excavator. 

Nummi and handmade pottery have parallels with the ephemeral character of short term settlement to the careful eye of the contemporary survey archaeologist. We know that local communities throughout history adopted flexible strategies to manage agricultural risk even during times of apparent economic, political, and social stability. During times of unrest or rapid change, like the middle decades of the 7th century, there would be a tendency to adopt more flexible approaches to survival and to shy away from longterm investments that would be more visible to the archaeologist 1500 years later. Like handmade pottery and nummi, ordinary features of everyday life would have persisted as low risk strategies and objects like imported pottery or large issue coins would decline as communities and individuals became less inclined toward significant investments or more substantial economic transactions warranting the use of larger coins.

The fragments of my writing over the course of a normal semester reflect the day-to-day strategies adopted to survive 21st century academic as a moderately productive scholar. The long, lazy writing days of spring break allow higher risk strategies to unfold, and these included interrogating intuitive connections and making obvious their relationships.

Coins, Raids, and Dates in 7th Century Cyprus

I plan to dedicate most of this week to putting words on the page for an article on settlement on Cyprus in the 7th and 8th centuries. The article will start with a brief treatment of the difficulties associated understanding settlement in this period. These difficulties which range from problems with the ceramic chronology to the dependence on poorly understood historical events to date archaeological evidence make the second half of the 7th century a particularly opaque period in the history of Cyprus.

The first issue that I’m working to tackle is what I’ve termed “the Heraclius problem”.   The emperor Heraclius (r. 610-641) yas always had a special place in the history of Cyprus. One of the first moves in the future emperor’s revolt was securing the island of Cyprus in 608 at which time he began to mint coins in his own name on the island. His efforts to secure Cyprus revealed an understanding of the island’s strategic significance for controlling the eastern Mediterranean. Heraclius’s appreciation of the island’s strategical value continued in his conduct of the Persian Wars. The island represented an important staging area for Roman forces cycling in and out of the east. The importance of the island as a base for campaigns in Syria, Egypt, and southern Asia Minor likely accounted for the interest shown by Arab forces in the middle decades of the 7th century and their eventual stationing of a garrison on Cyprus, probably on Paphos.

One result of the island’s special relationship with Heraclius and the key role that it played as a staging area for 7th century campaigns in the Levant is the ubiquity of coins minted during the reign of Heraclius. In fact, coins minted under Heraclius are the most common issues from the 5th-7th century on the island. This likely reflects the issuing of coins to pay troops moving back and forth through the island and the coins minted by Heraclius on the island from 608-610. Coins from Heraclius’s successor Constans II (r. 641-668) are almost as common. The number of coins drops precipitously after the reign of Constans II largely owing to the decline of regional mints and the political and economic ambiguity of the island as it passed into the strange condominium period during which both Arabs and Byzantines had some authority on the island. 

The abrupt drop in coins after the reign of Heraclius and Constans II poses an interesting problem for archaeology on the island. Because the number of coins declined so dramatically, we can probably assume that coins of Heraclius stayed in circulation for at least a generation, if not more. As a result, the use of coins of Heraclius to date archaeological features is a particular challenge. In the archaeology of Cyprus, however, this is a common occurrence. For example, coins of Heraclius date at least a dozen of the 70-odd Early Christian basilicas on the island. In most cases on Cyprus, the evidence from ceramics or other datable artifacts from stratigraphic contexts does not accompany the evidence from coins. Coins alone appear to date the structures. 

The use of coins to date buildings, destruction levels, and stratigraphy is problematic on archaeological grounds. Despite the appeal of coins as firmly dated artifacts, they are only useful if they are the latest object in a level. Moreover, the absence of coins from the 8th century on the island means that any dating by coins alone becomes problematic because of the uneven supply of currency to the island.

The use of coins to date these basilicas to the middle decades of the 7th century reinforces arguments for the collapse of Cypriot settlement and society at this time. In particular, the coins of Heraclius tend to support arguments from the destruction of these buildings as a result of the Arab raids on the island around 650. Of the 70-odd basilicas on the island, excavators have argued that a third of them were damaged or fell out of use over the second half of the 7th century and attributed this trend to the Arab raids. 

I sometimes joke with my ceramicist friends that our continued efforts to trace the use and production of traditional Late Roman red-slipped wares and storage amphora into the 8th century will eventual force us to change the dates on some well-known emperors. The “funny” is that I assume we can use ceramics to date coins and then to date political events just as coins have often been used to date ceramics. The ongoing revision of ceramic chronologies and a more critical treatment how coins work in an archaeological context are important steps in understanding 7th century settlement on the island.