Roman history is punctuated by agrarian crises and their associated political effects. However, new research challenges conventional interpretations of these rural developments and their broader significance. Late republican Italy is a good case study, being the subject of two well-known models, one general and one specific. The first argues that changing military conditions led to the breakdown of the seasonal integration of farming and warfare. As wars shifted further from Rome, it was impossible for soldiers to return to their farms to plough and harvest. Deprived of their principal source of labor, rural families fell into debt and were forced from their land, leading to a reduction in the number of free citizen-farmers over the last two centuries bc (Sall. lug. 41.7-8; Plut. TG 8.3; Brunt 1971). The Gracchan agrarian reforms were a partial response to this situation. A more specific issue concerns the long-term impact of the Hannibalic war on southern Italy: Toynbee’s famous thesis argues that the population and economy never recovered from the devastation, with repercussions extending to the modern era (App. BC 1.7; Cornell 1996; Toynbee 1965).
Various aspects of these narratives are now disputed. Close re-reading of texts suggests that soldiers were not seasonally discharged in time to return to their farms from as early as the fourth century bc; in other words, another means of reconciling war and agriculture already existed long before c.200 bc. The supposed incompatibility of military and agricultural duties after this date had therefore already been resolved and cannot explain the decline of small farmers. However, this raises the question of how war and agriculture were integrated if not seasonally. To answer this, Rosenstein (2004) creates hypothetical families of different sizes and age profiles and compares their calorific needs, their potential labor and the amount of land needed to sustain them. These families are then subject to various scenarios to assess their resilience in the face of military recruiting. The approach requires disparate data and questionable assumptions, and consequently the results should not be confused with reality. However, they place informative parameters on the claims of ancient and modern authors alike. The results suggest that small farms and warfare could indeed be compatible, and a “typical” family of five could spare the labor of a son. However, the critical factor was not when during the year his labor was removed, but when during the family cycle. By delaying marriage until the age of 30, men were recruited before they took over farms and started their own dependent families (on age of marriage, Scheidel, population and demography, section 4).
If war and agriculture were compatible, how do we account for the decline of small farmers during the late republic? Several scholars not only argue there was no decline to explain, but there might even have been substantial expansion. Without doubt, the Hannibalic war led to considerable short-term depopulation. But for surviving farmers, conditions improved: for example, there was automatically more land available; for surviving soldiers, marriage prospects also improved. Such conditions favor demographic growth (Erdkamp 1999). In fact, the census figures, though problematical, do indeed indicate growth. A key historical debate is whether this growth represents free farmers or the descendants of manumitted slaves. The latter has been the dominant view (Brunt 1971), though this is now disputed on several grounds. For example, the downward revision of slave numbers makes it impossible that their reproduction could account for the growth suggested by the census figures (Scheidel 2005). Although the late republic was a period of high population mobility (Scheidel 2004), immigration on the necessary scale seems unlikely. The best explanation is therefore the substantial growth of free rural populations. In turn, this radically different view of late republican rural demography requires reconsideration of wider issues such as colonization and the Gracchan crisis.
Such models are generalized across Italy. However, both historical and archaeological evidence indicate regional variation. In Etruria, post-war conditions promoted strong economic development. Largely untouched by Hannibal, the area experienced high demand for urban and military supplies, leading to inflated grain prices. Combined with immigration from the war-torn south, the local elite had incentive to intensify agricultural production and the labor and capital with which to achieve it (Erdkamp 1999). Archaeological evidence adds another level of detail. Close to Rome, there was a marked increase in site numbers, from the mid-second century BC through to a peak in the early imperial period. This area is notable for its density of settlement, including wealthy rural villas, and associated large population. In contrast, across inland Etruria the expansion of dispersed rural settlement peaked during the second century BC before a slight decline by the early imperial period. In this area, the density of settlement and population was much lower, and villas both fewer and less opulent (Witcher 2006). These contrasting patterns of settlement, agriculture, and demography suggest distinct late republican trajectories. The key difference was proximity to Rome, its growing market, and its wealthy elite (Morley 1996).
In Campania, military action devastated agriculture and population alike. But once Hannibal was expelled, Rome needed Campania to supply its armies. Migration and recruiting meant that labor was in short supply and survivors lacked the seed and capital with which to revive farming. This situation favored the use of slave-labor by external landowners with capital to invest (Erdkamp 1999). Again, archaeological evidence adds some detail. Military supply quickly gave way to specialized viticulture on large estates along the coast owned by Roman senators. Bulk wine was exported during the late republic, but declined due to provincial competition in the early imperial period. In contrast, inland areas experienced limited economic growth, but greater long-term stability (Arthur 1991). Again, external elite involvement and proximity to markets were critical to economic developments.
In southern Italy, the focus of Toynbee’s thesis, military action continued for longer, causing widespread devastation and depopulation. After the war, capital and labor were scarce, but by then falling grain prices meant there was little profit to be made through intensive agriculture in the short term. Investment and recovery were therefore limited and delayed (Erdkamp 1999). Archaeological evidence indicates significant settlement dislocation during the third century BC, some of it, perhaps, predating Hannibal (Lo Cascio and Storchi Marino 2001). However, though the dramatic and permanent decline of settlement around Metaponto fits Toynbee’s model well, it is exceptional (Carter 2005). Elsewhere the evidence points to slow revival, though there was much local variation. Particular trends include extensive colonization and villa building by absentee landlords developing slave-based viticulture on the coasts and sheep rearing and textile production across inland areas (Fentress 2005). Hannibal’s legacy was not permanent economic and demographic stagnation, but the alienation of landownership and subsequent vulnerability to external developments. In summary, the nature of the late republican countryside of Italy is still hotly debated; insights from comparative and archaeological studies mean new narratives are less “tidy” but more realistic.
World History
