Mac Gaw, Carlos García 2015, “The Slave Roman Economy and the Plantation System” (p. 77-111) in da Graca, Laura and Andrea Zingarelli (eds.) 2015, Studies on Pre-Capitalist Modes of Production. Brill.
Quoting the Roman historian Schiavone, Mac Gaw begins with the contention that “Most historians would agree” with Schiavone that “‘it is difficult to doubt that Roman society between the Punic wars and the century of the Antonini was a slave-based society.’” Yet, García Mac Gaw asks, what does a “slave-based” society mean?
Mac Gaw begins his investigation with a review of the major theoretical scholarship on slavery, noting Historians have long discussed the division between societies with slaves, and “slave-based” or “genuine slave societies.” García Mac Gaw notes how Finley tracked the interest in the subject in post-abolition society, and perceived a divide between moral and sociological approaches to studies of slavery. Finley’s characterization of the sociological approach is an economic one, as it came from authors he calls “economists,” and it privileges discussion of the inefficiency of slave labor over free labor, due to its higher costs (with a clear sympathy for the free market economy). García Mac Gaw goes on to try and explain this phenomenon, and poses that the issue stems from past studies of modern slavery which validated its existence as “a production factor of the [economic] enterprise, that is to say, the labour factor.” As such a factor, it became easy for authors to appropriate the “reasoning” of “economic rather than historical discourse to the analysis of slavery,” resulting in further studies that focused on the efficiency or lack thereof of slave labor, “especially in the United States of America.”
While a comparison of ancient versus modern forms of slavery is potentially helpful, once the challenges have been identified, it requires “extreme care by refining comparative methodology.” This chapter is an attempt to clarify such a method. There are, as Patterson noted and García Mac Gaw cites, sociological aspects that seem to reappear in different societies that practice slavery, enough to propose “some criteria for a unified approach for its study.” There are also economic aspects, and with both the sociological and the economic aspects in mind, it is risky to assume that both repeat themselves in different societies, “especially when comparing societies whose productive development is not equivalent.”
Historians have long discussed the division between societies with slaves, and “slave-based” or “genuine slave societies.” García Mac Gaw notes how Finley tracked the interest in the subject in post-abolition society, and perceived a divide between moral and sociological approaches to studies of slavery. Finley’s characterization of the sociological approach is an economic one, as it came from authors he calls “economists,” and it privileges discussion of the inefficiency of slave labor over free labor, due to its higher costs (with a clear sympathy for the free market economy). García Mac Gaw goes on to try and explain this phenomenon, and poses that the issue stems from past studies of modern slavery which validated its existence as “a production factor of the [economic] enterprise, that is to say, the labour factor.” As such a factor, it became easy for authors to appropriate the “reasoning” of “economic rather than historical discourse to the analysis of slavery,” resulting in further studies that focused on the efficiency or lack thereof of slave labor, “especially in the United States of America.”
In fact, Carandini observed that the interest in American slavery increased during the Civil Rights Movement, ending the “historical invisibility of Afro-Americans.” Fogel and Engerman’s Time on the Cross (1989) was one of the first works to question that free labor was more efficient than slave labor, but it remains one of many examples of how studies on ancient slavery have been dependent upon questions of economy with a comparative, modern capitalist context. In this way, ancient slavery – especially Roman – has often been explained in terms of modern plantation slavery, where the Roman villa has been studied as analogous (mutatis mutandis, as García Mac Gaw says) to the modern plantation, with the economic role of the latter in mind.
This analogy itself was born from the “comparative historical method,” which can be fruitful, García Mac Gaw adds, but carries lots of risks in this area of study. There is the particular problem of neglecting to compare the fundamental differences between ancient and modern “wage labour.” Furthermore, while a transhistorical slave mode of production seems possible, such approaches often neglect the contextual development of productive forces. Marx also noted this problem, at least according to Nippel, who says, “a closer look on Marx’s scattered remarks on ancient slavery reveals that they are almost all by-products of his interest in the American slave debate. [. . .] he [Marx] was remarkably reluctant to draw inferences from the modern evidence (as he understood it) with respect to the ancient world.” Thus, a comparison of ancient versus modern forms of slaver is potentially helpful, but once the challenges have been identified, it requires “extreme care by refining comparative methodology.”
Davis offers observations on looking at slavery from broader perspectives under the comparative historical method- the first is to study the history of human bondage, the second is to study the historical forces that led to “racial slavery,” and to understand this in the narrative of the costs of ‘modernization.’ Davis also advocates for careful empirical comparison and acknowledgement of “the Big Picture” of interrelationships that “
Constituted an Atlantic Slave System as well as the place of such racial slavery in the evolution of the Western and modern worlds.” Dal Lago and Katsari noted a desire to develop global or comparative historical studies of “slave systems.” They note that a rigorous approach is often needed, but a “soft” approach is often taken, with only hints of rigor, and only a brief treatment of significant themes.
From these approaches, the comparative method appears limited when compared to a global approach. The apparent “broadening” of comparative scope, and thus potential, in the global method is problematic, because it does not develop from the comparisons established, but instead tracks the “inner logics that structure the object of the study according to the dynamics of its functioning, what Davis calls ‘the Big Picture’.” The first comparative approach is more sociological in its nature- it looks for sociological types, such as where slavery exists in different contexts. Issues arise in the disparate circumstances of historical context, where evidence is limited and their are important factors such as sex ratios to take into consideration. Such problems are resolved, under the global approach, by employing the logic of one system [slave plantation, for example] to explain/fill in the gaps in the comparative example [ancient slavery]:
“Such aspects include, for insance, issues that relate to slavery as an answer to the dynamics characteristic of the European capitalist market and the American and African integration in that economic framework. The notion of labour market in this case may explain the development of slave exploitation as a response to the lack of labour force in a space where it is not feasible to control the labour force due to the availability of alternative meas of production (free lands). [. . .] Accordingly, the application of the comparative method should be extremely cautious insofar as it affects the underlying logic that explains the dynamics of the social structures. All these precautions will not be present if the model of modern slavery works as a non-explicit assumption for the study of ancient slavery, and the model of the capitalist plantation imposes its logic on the functioning of the Roman villa. In this sense, the use of the economist perspective to organise the view on ancient slavery appears as the most persistent epistemological obstacle to guard against.”
The Income Composition of the Landowning Class
As Bradley’s comments on when Rome became a slave-based society show, there is a paradox in answering this question, as his historiography proposes “the existence of a slave-based society before it really came to be ‘a genuine slave society’.” To account for this, is it worthy to ask why Rome has been considered at all as a “slave-based society.” For Hopkins, the moment occurred in Roman history when 20% of production was accomplished by slaves. Finley was a critic of the ‘numbers game,’ as he called it, because the evidence does not exist to quantify slave productivity in this way. Garcia Mac Gaw believes that the ratio of slave labor to free labor is important, but the numbers themselves are not what is at the center of these arguments. It is rather the qualification of “slave-based,” without a center on the economy, that such numbers get used for; furthermore, they avoid the topic of social relations of production.
What other ways have been used to define what constitutes Rome as a slave-based society? Finley proposed that two issues be addressed, which I pose as questions: one, who are the slaves’ owners? And two, what role do they [the slaves] play in the economy? Dal Lago and Katsari offer separate criteria, beginning with a definition that first assumes ‘slave-based system’ refers to a society where slavery sustains the economy, and also where “(a) elite wealth and slave ownership were two notions inextricably connected to each other; (b) a large part of the trade revolved around buying and selling slaves; (c) a high percentage of the workers were enslaved labourers; and/or (d) states and other types of institutions relied on the profits made with slavery for their prosperity.” They also distinguish the ‘slave society’ – which holds the economic basis – with the “slave system,” which refers to the spread of the institution of slavery- “an instution based on the “slave mode of production” and system of labour.”
If we consider the third letter of Del Lago and Katsari’s criteria, that “(c) a high percentage of workers are enslaved labourers,” it is necessary to prove that is the case. Finley does so without numbers, but explains that large productive units either had slaves or “copied the peasant domestic unit.” Ste. Croix proposes that the “owner class” got their surplus from “unfree labour” with chattel slavery as one example of that unfree labor. Yet, Ste. Croix still calls the ancient world a “slave economy.”
In the face of these qualitative conundrums, where does one even begin to try and get the numbers necessary to justify the claim? Garcia Mac Gaw proposes that there are three questions to answer:
- What percentage of the total agricultural production corresponds to slave labour power in relation to free peasant labour power in the Roman Empire?
- What percentage corresponds to coloni rent in relation to the slaves in large estates?
- What percentage of landowners comprising the elite obtain the largest part of their income from slave labour in relation with the rent provided by the coloni?
“Regarding point 1,” he says, “many historians agree that the bulk of the product of the Roman Empire was obtained from the labour of peasants with smallholdings.” Those peasants in many cases also had slaves, and their contribution to these “domestic units” and “small workshops” would need to be accounted for. On point 2, Garcia Mac Gaw says slave rent is impossible to calculate in the “whole agrarian income for large domains.” Furthermore, the slaves of the coloni would also need to be accounted for, since they “contributed to the tenants’ labour.” The size of the plot and nature of the labor would also produce different outcomes for quantifying the slave labor in relation to coloni: a small plot would be “organised outside the logic of the large plantation,” and thus “reproduced the domestic structure.” If the plot was great and only slave labor was used, then the plantation model works. The other option is that the “great coloni” mixed slave labor with sub-tenancy for small coloni, and the domestic structure was thence repeated.
Point 3 assumes that the bulk of productive products comes from lands in Italy and Sicily, but it is historically the case that many landowners had properties in other provinces that were also productive and run by coloni and “to a lesser extent by slaves.” The confusion seems to arise from the restrictions on wealth accumulation imposed on the senatorial classes, as Rawson describes: “the position was of course entirely different under the Principate; by Seneca’s time the typical rich senator had estates all over the Mediterranean world, though he still needed permission to leave Italy. In the Republican era there was naturally no restriction on travel or ownership abroad for equites or humbler persons.”
Free Peasants and Coloni
García Mac Gaw next draws our attention to the issue of how the income of the “landowning class” was composed to question “the prevailing historiographical views on slavery” (he also acknowledges that it is not the only factor of interest). From the traditional historiographical narrative, slavery developed in response to depopulation in central Italy and “the crisis of the small owner peasant sector.” García Mac Gaw cites Hopkins, De Neeve, Kolendo, and Sirago as comrades in this view, but presents differing opinions from Garnsey and Frank on the state of the “small owner peasant class,” all of whose studies suggest that there was some support for this class at the time through various policies. Harris, meanwhile, challenges the depopulation narrative with the fact that Sulla resettled some 120,000 soldiers in Etruria on lands confiscated from people who opposed him. Harris also cites the continued foundation of colonies under the Second Triumvirate.
García Mac Gaw notes that such places were occupied by individuals with varying degrees of wealth, citing the praedium Arretinum of Atticus and the “properties of Domitius Ahenobarbus in Cosa and Igilium,” both of which could have been obtained and regifted from Sullan occupation. Both Etruria and Campania seem to have been “affected by triumviral and Augustan colonisation, perhaps due to the existence of coveted lands in those areas.” Rich areas in Capua and Campania, especially the ager Falernus (known for its wine) and peninsula of Surrentum, were probably worked by small peasants and rich absentee owners, as Carrington (1931) noted (and with whom García Mac Gaw agrees).
Part of the problem of studying these properties is that there is documentation of different scales of land owned and worked, but little archaeological evidence to shed light on smaller properties. Thus, the narrative that small farmers declined, and large properties increased in the Late Republic is difficult to measure from the material evidence. Dyson (2003) noted that archaeological research produced different answers on this front due to, as García Mac Gaw says, “methodological issues and to ideological preconceptions.” Liverani seems to have been one of these, as one who measured a reduction in archaeological sites from the third to second centuries BCE in the same regions studied by Dyson (ager Cosanus and the ager Veientanus), but did not have “selective criteria in favour of large properties against small ones.”
Intensive occupation of the ager Veientanus in Etruria occurred during the Augustan age and lasted to the second century CE. Liverani praised this as the success of the villa during the Principate. “However,” García Mac Gaw says, “while [according to Liverani] this implies a greater use of slave labour in teh southern area of ager veientanus, it also implies a greater use of colonate in the northern regions. According to this author [Liverani], there are no grounds for the claim that the disappearance of the small peasant property can be attributed to the consolidation of the villa or, even more inaccurately for this period, to the system of large estates.”
Subsequent studies – including new archaeological evidence – have shown that more small farms must have existed in the Roman countryside than previously thought, challenging the old narrative of extreme depopulation in these areas. The implications challenge the other old narrative that the Italian peninsula became a slave-based system as a result of the success and domination of the slave-based villa. Instead, the coexistence of diverse forms of land ownership has emerged.
The coexistence of diverse forms of land ownership was accompanied by diverse forms of land management. García Mac Gaw points out that the colonate was the most widespread form of labor exploitation in the territory of the Empire, and dating back to the Late Republic. He offers support from Caesar, Sulla, Cato, Varro, and even Horace, highlighting how these authors give diverse forms of land management, from coloni to free tenants, to vilici, mercenarii (or obaerarii), and politores. Slaves, like the vilicus, could oversee land in similar ways to free tenants; likewise, free men could be hired to take on tasks also performed by slaves.
García Mac Gaw goes on to say, citing the work of Capogrossi Colognesi, that while absenteeism was widespread amongst the “great landowners,” this occurred simultaneously with agronomists who were present on their lands. In the former case, but perhaps not restrictively, the property was run by vilici and slaves. There is no reason to suppose that absenteeism, even if it was the standard, should imply that leasing land to free tenants should be any less present at this time. In fact, references to coloni in Pliny suggest that there are different peasants leasing his land that he has to (begrudgingly) deal with.
Forms of Slave Rent
Some slaves were servi quasi coloni, that is, slaves who operated like free tenants, “who were given a plot of land to work in exchange for a rent payment.” The surplus produced did not form part of their peculium, or at least was not guaranteed as part of it, as legal cases from the Digest illustrate. Likewise, in one case, one slave made ten thousand sesterces as a servus quasi colonus, which could count in theory count towards the price of his freedom. Instead, the jurist ruled that the money belonged to the heir not “for his freedom but for his labour (as salary: pro operis).
In the rest of this section, García Mac Gaw seeks out other examples to prove how these slaves were not a “missing link” between chattel slavery and the colonate, but an alternative form that existed alongside both. The servi quasi coloni were born from locatio conductio with free coloni– hence, the complications in certain legal cases arise because both exist, and the circumstances of the servus require special attention in different roles- particularly where the servus acts as both res and persona. There is a great emphasis in the legal cases on how the profits of the servus are to be allotted, and in many cases they did not belong to the servi as part of their peculium, but instead were for their masters for the “services” (operae) they performed. “[W]hat is noteworthy is the plasticity allowed in order to obtain the surplus produced by slaves rather than the rigidity that suits the model.”
All of which brings García Mac Gaw to question, again, the assumptions made when discussing the “slave based system” of Rome in this context, and he notes that “there is not a particular mode of appropriation of the surplus of slave labour defining this mode of production.” The primacy of ergastulum, or “gang slavery,” only comes from historiographical narratives that have erroneously seen ancient slavery through the singular lense of modern plantation slavery. “In the ‘plantation system’ the slave profit is expressed as labour. [. . .] [T]herefore all the slave activity in the proprietor’s land is equivalent to surplus value. [. . .] In the case that the slave acts independently by renting himself out, the master’s surplus value will be equivalent to the money the slave pays him, and the slave is in charge of the reproduction of his own slave labour power. This last example is identical to the case of the servus quasi colonus: in this case the slave pays his master a money-rent or a rent in kind that is identical to the surplus value and the master is not involved in the reproduction of the slave labour power.”
To better explain – or rather, reconfigure ancient slavery in terms of modes of production, García Mac Gaw turns to the tributary mode, rather than the slavery-based mode. Part of the reason is that there is an economic implication in labeling ancient society as “slavery-based” that has enough overlap with legal criteria to appear less “genuine” than it seems. García Mac Gaw believes that the evidence shows that slavery was not a necessary form of surplus collection, nor necessary form for the functioning of the Roman economy. Rather, it was more of “a legal status without direct economic consequence.” Speaking hypothetically, he says, “Therefore, the slaves [if defined by legal staus, rather than economic] would be exploited in different ways, and as a result there would not be a single slave mode of production but several, among which we could find, obviously, the slave mode of production of the plantation. I tend to agree with this second notion.”
Scope of the System and Dominant Class
Here García Mac Gaw gives a brief comparison of nineteenth century examples of slavery. He presents Engerman and Genovese’s analysis of a study done by Martins-Filho and Martins on the slave-based economies of Minas Gerais, Brazil, and Virginia. E. and G. noticed that M-F. and M. do not rebut the assumption that slavery was required for plantations existing for exportation, or vice versa, in their examples. However, E. and G. note that in slavery persisted in Minas Gerais and Virginia for reasons that did not make economic sense- so the question is why slavery nevertheless persisted? “According to Engerman and Genovese, what the study they review shows is not that the slave-based system can exist without the world market, but that the slave-based system –in this case the Brazilian one as a whole, which was dependent on the world market –could expel and support economic subsystems based on slavery, though isolated from the market sector.” That is, E. and G. propose that slave-based economies of the nineteenth century show a consistent implication of those societies within the capitalist framework, and thus “could not generate an alternate slave mode of production reminiscent of that of the ancient world.”
García Mac Gaw disagrees- at least to the extent that E. and G. imply that the ancient slave-based system did not depend on “its insertion in a greater economic system.” The variable forms of slave labor exploitation, which García Mac Gaw bins under the “ancient tributary mode of production,” were only made possible through the “huge capacity of economic accumulation due to the fast expansion throughout the Mediterranean.” The biggest different is that the slave-based ancient economy was not necessary “for the elaboration of commodities.” Expansion did, however, produce and affect forms of taxation.
Yet, was the “dominant class” a “slave-holding class”? And how does one explain the “crisis” of the slave system? For the first, García Mac Gaw says that the answer is “yes,” but this same class was also made up of those who were not slave-holding landowners. “Indeed, whether a Roman landowner exploited slaves or tenants did not affect his social position. To the second, the “crisis” came about from a “withdrawal or abandonment of direct exploitation through gange slavery,” and the same “crisis” is nullified if we accept that Rome did not have a slave-based economy. “Most likely, the progressive disappearance of the plantation system should be seen simply in relation to the supply for the trade circuit linked to the functioning of the rent originated from tribute.” And thus,
If the ‘slave system’ is a form among others of exploiting slaves, the modification of such forms does not make the class that appropriates slave labour more or less of a slave-holding class. Otherwise we would fall into the paradox of labelling the landowners who organised their land exploitation under the plantation system as ‘slave-holding’; but these same landowners would not be ‘slave-holding’ if their slaves were exploited in a manner similar to that of free tenant farmers. Mac Gaw offers two conclusions:
- “In societies where the institution of slavery has developed deeply, it has been dependent on other dominant social-economic modes existing in those societies.”
- “In pre-capitalist societies, the mechanisms on which surplus appropriation is organised by the dominant classes are of an extra-economic nature. In Roman society, slavery only reinforces the position held by the class of great landowners in relation to the control of the state apparatus. The slave labour system of the plantation is not a determining factor, aside from its economic consequences, but the control over the lands is a determining factor, as it reinforces the mainly political role of the roman landowning class in relation to the city-state structures through the villa system.”
Both are at odds with Marx’s own conception of slavery and fundamentally undermine comparisons he sought to make with modern wage-labor.
Leave a Reply