In an important prolegomenon to her new book, Surviving Rome, Kim Bowes calls for a qualitative, microhistorical approach to Roman economic history, in order to build the large mosaic of the Roman economy from smaller, case-based tesserae.Footnote ¹ Such a bottom-up approach would contrast the large-scale, muscular models like those of Scheidel and Friesen, which take on the whole of the Roman economy, with more nuanced accounts of its separate parts.Footnote ² Paolo Tedesco’s new book offers just such a microhistory: a detailed analysis of the world of the Albertini tablets, produced by the peasants who lived on the Fundus Tuletianos in southern Numidia at the end of the 5th c. CE. He then uses the vantage point established by this microhistory to examine a much longer chronology of the relationship between producers, landowners, and the state.

The book begins with an overview of the settlement patterns of later Roman North Africa, starting with the extensive imperial properties, which were leased to members of the elite or farmed directly by conductores through tenants. The emperor owned between a quarter and a third of all the arable land, while most of the rest consisted of large, privately owned estates, the patrimony of Roman senators such as Symmachus, the Roman churches, and the provincial élites. Further down the scale, the curial class, and with them the African bishops, were always landowners. There is not much trace of smallholders; although surveys such as that of Dougga show a dense network of occupied sites, these may have been the homes of tenants rather than independent landowners.Footnote ³ At the bottom rank of the producers were the wage laborers, who worked seasonally, perhaps to supplement inadequate incomes from their own farming. The African picture, as elsewhere, is thus a highly variegated one, but as a whole seems to consist largely of peasants bound by more or less onerous contracts of tenancy to landowners of various ranks.

Over all loomed the demand of the Roman state for taxes, and specifically the annona that fed Rome. Tax collection was the responsibility of the municipal curiales, who were either oppressed by shortfalls or enriched by skimming, depending on their networks and skill. This fiscal role of the municipia has been stressed by Chris Wickham as fundamental to the prosperity of the towns.Footnote ⁴ The arrival of the Vandals changed some of that, and broke the link by which the grain of the annona traveled to Rome. Gone, too, were the municipal obligations: in all instances adduced by T. we see the peasants paying their taxes through the landowners, who seem to have paid them directly to the tax collectors, eliminating the curial role and, with it, the structure of the municipia. With the disappearance of the annona, the total tax burden may have diminished as well. This may have led the most marginal peasants to revert to subsistence agriculture, although on the coast market production and demography saw little change: where once wheat and oil went to Rome via taxation, we now see them as cash crops, enriching, in particular, the Vandal kings.

The heart of the book lies in the next two chapters, the discussion of the Albertini tablets. This archive consists of 45 wooden tablets with Latin texts in ink, found in an amphora somewhere on the border between Tunisia and Algeria.Footnote ⁵ Long the only such archive known, it has now been joined by the even richer but presently inaccessible Sfez archive, probably found in the area of Musti, also written on wood.Footnote ⁶ The legal documents these archives contain include bills of sale, lists of dowry items, wills, and leases. The interpretation of the Albertini tablets, treated first by Eugène Albertini, then by Christian Courtois, Louis Leschi, Charles Perrat, and Charles Saumagne, and further discussed by Jérôme Carcopino, has until now gone unchallenged.Footnote ⁷ For these scholars, the archive appeared to show a steady decline in the conditions of the local peasantry, forced by looming crisis to sell the leases to their tiny parcels of land to local landowners as their poverty increased and they reverted to a subsistence economy. Striking is the continuity of the particular emphyteutic tenancies known as Mancian, first attested in the Lex Manciana of the early 2nd c. in the Mejerda valley.Footnote ⁸ These tenancies appear to have combined a series of local customs, including specific sharecropping arrangements and corvée labor. They permitted peasants to lay claim to unoccupied land by planting trees or vines, and allowed them to remain in occupation and bequeath the lease to their children as long as the agriculture was not neglected.

T. offers a detailed picture of the community of the fundus, based on peasant households whose members might rent parcels of land from the local estate while also farming as smallholders in their own right. His analysis identifies five other major landowners besides the proprietors of the Fundus Tuletianos, the Geminii, whose archives these appear to be. There are also upwards of 250 peasants of various standings, from smallholders to tenants, and any mix of the two. Smallholders could rent parcels from the fundus as their family size increased, and with it the available labor. Conversely, as sons married and moved away, leases could be renounced. Nomadic pastoralists complemented the agriculturalists, providing labor at harvest and manure for the stubble their flocks grazed on. Slaves (only one is recorded) would have simply supplemented the labor of the peasant families, or provided domestic service to the landowners. The landscape was rugged: with less than 200 mm of rainfall, agriculture depended on irrigation and the skillful management of water, which in turn required investment of both labor and capital. Subsistence strategies clearly favored mixed cultivation of trees, grain, and vegetables, with olive oil and perhaps dried figs and almonds the only surpluses entering the market. The fundus, however, was far from free of fiscal pressure. To the state were owed both land tax and the head tax, the capitatio, the latter paid by the peasants themselves. Here the landowners acted as intermediaries with interests that diverged from both the state and the peasants. Literacy and basic accounting were key aspects of this role. A dominus such as Geminius Catullinus served as a crucial link between the peasants and the state, using his local knowledge to extract the sums due. T. draws a clear distinction between the tributary obligations of the peasants, who owed the landlords both rent and taxes (probably in kind), their moral economy, which prioritized their survival and reproduction, and their direct participation in the market, largely through a surplus of olive oil. Of these categories, the most interesting is the moral economy, which comprised a whole series of customary rights and obligations rooted in traditions that may have been pre-Roman in origin. Chief among these would be the individual’s obligation to the survival of his own family, both nuclear and extended. We can also see these customary obligations at work in the witnessing of documents by more prominent and trusted members of the community.

The fourth chapter is a deep dive into the documents, beginning with a demonstrative attempt to map several clusters of them by plotting the position of a given sale in relation to its neighbors. The plots sold are generally small, containing as little as a single olive or almond tree. There follow discussions of the buyers, sellers, scribes, witnesses, and those who owned land outside the estate, including another Geminius, Ianuarius, brother of Felix, and Cresconius, whose activities are prominent in the archive though he lived outside the fundus. He was clearly a substantial farmer, whose daughter inherited a significant dowry, which, as she moved away, then disappeared from the estate. His two sons served as witnesses. Typically, the daughter’s dowry did not include any land: in pre-modern Kabylie, daughters inherited only buildings in the village, while agricultural land and trees passed to sons alone.

At the bottom of the ladder were those who exchanged their labor for brief access to land – a form of payment which is striking, but apparently not unusual. These individuals tended to bear non-Roman names and probably supplemented the estate’s workers at harvest and in tending livestock. As the well-known inscription of the “harvester of Mactar” shows, none of these conditions should be read as permanent:Footnote ⁹ T. insists on the fluidity of the peasant community, the expansion and contraction of families and with them their labor power and the size of their holdings. Far from being a unilinear contraction in peasant holdings, T. shows that the choices of the peasant household as to what it cultivated were made in reaction to changes in the peasant household itself. The peasant’s agency, while constrained by the owners, the state, and the vagaries of the harvest, was always part of the story, enabling the sons of illiterate fathers to acquire land and to learn to read, becoming respected members of the community. T. sums this up in an optimistic passage:

However implausible the last element of that formula, it is not directly contradicted by the Albertini tablets. What is striking, however, is that, but for a single presbyter, there is no trace of the church here – no nuns, bishops, monasteries, or churches. This seems a rather different world from that treated by Leslie Dossey, who describes how, over a 150 years earlier, Christian teaching and its diffusion in the countryside by the clergy effectively empowered peasants to resist more forcibly the multiple impositions of the powerful.Footnote ¹⁰ The picture painted of the Fundus Tuletianos on the basis of the tablets includes no debt-collectors, no circumcellions, no peasant resistance, and above all no churches or ecclesiastical properties. One of the landowners is even described as a flamen perpetuus, one of the last recorded uses of the term.Footnote ¹¹ The political and social power of the landlords over their tenants appears to stand alone, bolstered by their tax-collecting role and not moderated by the church. The opportunities this gave them for exploitation might, however, have been mitigated by T.’s second economy, the customary rights and obligations of the moral economy.

What is striking about this detailed discussion is that the tablets themselves give us no idea of the settlement patterns in the area. Did the Geminii live in elaborate villas? Residence in a nearby town is hardly likely, as there aren’t any nearby towns. The closest, Theveste, is almost 100 km away. And, crucially, did the peasants live in a village, or in scattered farmhouses? We should note that no plots are recorded as containing domestic structures. Wilson’s analysis of the probable find-spot points to a large site, Henchir ed-Debine (AAA 51.60), measuring perhaps 8 ha and possibly walled, on the southern slope of the Jebel Mrata around 100 km south of Theveste.Footnote ¹² This settlement could have included both the dwellings of the Geminii family and housing for the peasants of the tablets, who would have farmed plots in the neighboring valleys and on the piedmont plain. That there is no probable toponym for this site in the tablets is striking, though – many of the plots must have bordered on it. Wilson’s examination of aerial photographs also reveals two probable centenarii, or fortified granaries, but no irrigation channels whatsoever. Wilson explains their absence, and the term aquaria, repeated multiple times in the tablets, as referring to irrigation by wells, the aquaria being water-lifting devices known as rotae, which brought water up and dropped it into shallow channels leading to individual trees. The Arabic bir, “well,” remains a significant toponym in the area. In any case, if the bulk of the peasants simply lived together on the estate of the dominus, this would have fostered the community of interests and network of relationships that the archive reflects.

For the next two centuries we have far less information, although T.’s contention that in the old Roman provinces life under the Byzantine state continued much as it did under Vandal rule – if on a diminished scale – is at least questionable. There was, of course, no revival of the fiscal duties of the municipal elites: traces of these elites are vanishingly small, as they were replaced by high-ranking soldiers, probably absentee landlords. Landowners and landed peasants alike may have continued to pay capitatio and iugatio, collection organized through the landlords. Our evidence for peasant participation comes largely from the construction of fortifications, like that of Aïn el-Ksar, which records 19 “citizens of this place” who built a castrum on the orders of the magister militum.Footnote ¹³ Why this demonstrates continuity of the tax structure, as T. suggests, is not entirely clear: the interpretation appears to hinge on the identification of one of the citizens as a magister, which could mean either the person responsible for the imperial estate on which the fortification was built, or simply the representative of the community. The fact that his is the one Greek name on the list favors the former interpretation, and might suggest continuity in the management of imperial estates. T. takes this as evidence for a form of conscription based on the iugatio, whereby instead of money the landholder was required to furnish soldiers. The shift in the currency of taxation from money or produce to labor or military service is striking. Though labor had always been a form of exchange, the conscription of peasants as builders or soldiers would have resulted in considerable decline in agricultural production and a disinclination to continue farming. Nor is it clear that the castra replaced villages in the settlement hierarchy, rather than simply forming part of the strategic networks of the state. The disappearance of peasants to military service would have further weakened what remained of the local elite, of which there is now very little trace, particularly on the desert margins. The problem became acute by 570, when a law was promulgated to ensure that the children of free women and tied coloni had to remain on the estates. Although T. is clear that settlement was in decline in many areas by the late 6th c., he offers an optimistic picture of the period, concentrating on Zeugetania, where he sees a government shift in strategy from requisitions or taxation to remunerative purchases that would have enriched the owners of great estates on the coast and in the Mejerda valley. The excavation of an oil-producing farm near Thugga, completely rebuilt at the end of the 6th c. and occupied into the early 8th c., illustrates how prosperous these sites could be.Footnote ¹⁴

His optimistic reading of the Aurès, based on aerial photographs, is far less convincing (p. 225). This chapter is notably lacking in any serious engagement with the attacks from the Sahara, certainly present in the 530s but which it can be argued, both from archaeological and textual sources, date as early as the first quarter of the 5th c.Footnote ¹⁵ The presence of an “imperator,” Masties, in the Aurès in the 5th c. hardly suggests that the Vandals or Byzantines had any means of tax collection there,Footnote ¹⁶ while the massive Djedar tombs west of the Hodna basin, again from the 5th c. onwards, are not those of small local landlords, nor of Vandal or Byzantine officials. It is from this period that we see a new branch of Tamazight (Berber), spread across North Africa from the Aurès to the Rif, suggesting an influx of new populations.Footnote ¹⁷ The lines of Byzantine fortifications, whether small towers or forts, do not speak to a peaceful, productive economy, but to one increasingly constrained by the need for defense. Admittedly, on the coast from Carthage through Byzacena and in the Mejerda valley there is no evidence for disruption, but the fate of the peasants elsewhere is far from established. Both the Kasserine and the Jerba surveys show a massive decline in settlement. Whether the peasants moved to the coast and the cities, or met the fate of those described by Augustine, drained from areas under barbarian attack and enslaved, is impossible to know;Footnote ¹⁸ continuity of cultivation on the margins is very hard to argue for. At this point it would have been useful to know precisely where in North Africa the widely attested seals of the kommerkiarioi, representing the imperial government’s management of resources for the army, have been found.

As T. points out at the beginning of the final chapter, we know even less about the subsequent two and a half centuries, from the first Arab raids in 648 through to the end of the 9th c. Though there may have been some collaboration with the remaining local elites regarding taxation, there were certainly land confiscations, while in spite of their military service or conversion, the Berbers were subject to enslavement, with their bodies used to pay taxes. The revolt of 739–740, led by the Khārijite sect, was suppressed some 20 years later, but left much of the West independent, with small city-states like Tihert, Tlemcen, and Walīla (the former Volubilis) developing their own political and economic structures and coinages. While the estates of Ifriqiya remained in the hands of the conquerors, much of the West was lost. T. rightly dismisses Talbi’s argument that these large estates were run as plantations worked by slaves, an inheritance of late Roman practice. Never widely accepted, the near-absence of slaves as a distinctive component of late Roman North Africa makes it difficult to understand why the Arabs would have adopted the practice, and it is equally unclear why slavery would be useful for cultivating wheat and olives. Ifriqiya is not the marshes of Basra. Instead, peasants could rent land for cash, or participate in various sharecropping arrangements, down to the khimāsa, by which the worker received a bare fifth of the crop – although our earliest evidence for this arrangement dates only from the 11th c.

In his conclusions, T. returns to the themes of the complex stratification of the landowning class and the variety of contractual arrangements by which land could be worked. Individual tenants might pass through many of these over the course of their lives, either beginning or ending as landowners. As Vandals took the place of the Roman state, to be replaced in turn by the Byzantine state and eventually by the Arabs, there is a steady decline in production and a contraction toward northern Ifriqiya, but until the Arab conquest nothing substantial changed in these fundamental relationships. We see, perhaps, an abatement in olive oil production as international trade faltered, but it was certainly still a motor of the economy at the time of the conquest.

This is all very convincing, although it leaves a good deal out. The Albertini tablets appear to reflect an entirely normal set of Roman property relations, however complex. But like the church and the relationships it engendered, tribes – so much a part of late Roman Africa – are almost entirely missing from the picture. While the family of Nubel, including the fascinating figures of Firmus and Gildo, is mentioned, the implications of the tribal relationships visible in the history of that family, who code-switched between local princes and Roman officials, are not examined.Footnote ¹⁹ How do tribal relationships modify the “moral economy”? Do peasants pay tribute to their chief? Is there redistribution? What is the relationship between the tribal lands and the state?Footnote ²⁰ Most of all, however, we lack the materiality of the lives of these peasants. Granted that no small late Roman farm has, to my knowledge, ever been excavated in North Africa, one can learn a great deal from surveys such as that of Dougga.Footnote ²¹ Missing materials must also be taken into account. T. barely mentions textiles, which imply relationships with herders as well as markets, and must have been an important part of the economy in the world of the Fundus Tuletianos; even the dowry of Geminia Januarilla contained a loom.Footnote ²² Textile production, probably carried on largely by women, would have provided a source of household income as well as supplying the basic needs of its members: its invisibility in almost all of our sources, both textual and archaeological, creates a major gap in our understanding of the economy of the peasant household, the strategic deployment of its labor, and the relationship between the peasants and pastoralists.

How the peasants lived – whether in villages or dispersed farms – matters as well, for their social structure, their agency, and their marriage patterns. As regards pottery, only finewares are mentioned. While it is true that the production of African Red Slip ware (ARS) continued to the end of the 7th c. and even into the 8th c., it is less important to know that it reached Rome than to know whether it reached, say, Sbeitla. The Dougga survey has shown that many rural sites remained occupied but were no longer receiving ARS. There is little doubt that the market economy was by this period strictly regional, if not completely local: in the 4th and 5th c. we can already trace local ARS productions – at Sitifis, Castellum Tidditanorum, Diana Veteranorum, and Thamugadi – that appear to have been marketed within a radius of around 35 km.Footnote ²³ In Numidia by the 6th c., no finewares at all are found on rural sites near Diana Veteranorum: the landscape appears entirely deserted, although perhaps a deeper study of the coarsewares could fill the gap. What is clear from the survey of the territory of Diana was that the settlements highest in the hills had the highest proportion of handmade pottery, again showing the engagement of women in household production; the anthropology of the 19th-c. Kabylie shows that handmade pottery was exclusively the work of women. If the area was occupied at all in the 6th c., what little agriculture existed was clearly reduced to subsistence. We are thus not dealing with all of North Africa, only those parts that still paid their taxes.

Although all of these aspects could have been developed further, the book is nonetheless a very valuable contribution, offering a fresh reading of an important set of documents supported by an enormous bibliography. It is well written and produced, though at the very high price, the editor might have provided a copy editor.Footnote ²⁴ More illustrations would have been welcome: these are limited to elegant and generally accurate maps.Footnote ²⁵ However, T.’s focus on the strategies of reproduction of the peasant household is a crucial one, even if enormous gaps in our knowledge remain.