Return of the Slave Society

Aristotle dreamed of the robot revolution. A slave is a living tool that serves multiple purposes; likewise a craftsman’s assistant (Politics 1253b). This is demonstrated by the fact that, if every tool could perform its own work when ordered, or by seeing what to do in advance, like the statues of Daedalus or the self-moving tripods of Hephaestus, craftsmen would have no need of assistants or masters of slaves. Tools are an essential component of the state; workers, maybe not so much.

Back in the 1980s, the Italian archaeologist Andrea Carandini sought to characterise the impact of the rise of the ‘slave mode of production’ in Roman Italy by describing slaves as “ancient computer-robots”. The point was that the transformation of the Italian countryside in the 2nd-1st centuries BCE was not just a matter of the replacement of one sort of worker with another (free peasants with slaves), or of the reorganisation of the land into larger, more productive units, but of the introduction of ‘thinking tools’ that could be set complex tasks within a new semi-industrialised production system. The consequences of this change were then seen in growing inequality, political instability, the anger of the now displaced and impoverished peasants etc., leading to the collapse of the old political order – but not to any changes in this system of exploitation.

I always found Carandini’s argument disturbing because of the risk of complicity with the slave-owners’ perspective, seeing slaves solely in terms of productivity and erasing their humanity (I was at the same time reading similar criticisms of Fogel and Engerman’s Time on the Cross, which attempted a cliometric analysis of the economic benefits of slavery in the USA). But it continues to offer food for thought; and, in the face of a continuing stream of articles and television programmes about the rise of the robots and the inexorable spread of automation in the present, I’ve started to wonder about the possibility of focusing the analogy in the other direction: what can we learn about our possible future trajectory from studying a past society where ‘autonomous tools’ were ubiquitous?

Automation permeated the Roman economy; it’s not that every enterprise employed robots, but the largest, most market-oriented ones did, vastly increasing the profit margins and market share of those who could afford them. Moreover, wealthy Romans handed over much if not all of their financial affairs to the machines, which were theoretically dedicated to the service of their masters but enjoyed significant autonomy and independence (assuming that their masters could have understood their dealings if they troubled to investigate).

Even more significant for our purposes is the way that non-human intelligences could be found in every area of social life as well, to the point where their supposed owners ceased to register their presence much of the time. Roman masters entrusted confidential information and secret messages to their machines – and lived their lives surrounded by them, carrying out all the mundane tasks of household management and personal care in the background, silently watching and learning.

With high rates of divorce and remarriage, as Keith Bradley has argued, children might be largely raised by robots, forming their most long-lasting and intimate bonds with their robot nannies and tutors. Some adults, too, indulged in the pretence of equal, affectionate relationships with certain of their non-human servitors, perhaps even believing that their responses were genuine and spontaneous (“Thank you, Siri.” “You don’t have to thank me.”). But they equally regarded them as unproblematically available for sexual exploitation or physical abuse; they were things, not humans, even if they sometimes looked and behaved like humans; they were property, to be used as their owner wished.

Not implausible? The aim of the exercise is not to make us think differently about Rome, but to establish a basis for developing the analogy for the modern period, that the pervasive presence of artificial intelligences in our lives is sufficiently similar to the pervasive presence of slavery in the Romans’ to make it worthwhile exploring further. Because a crucial point about a ‘slave society’, as discussed by historians like Finley and Bradley, is that it affects (infects?) everything. Social relationships, family relationships, economic activity, cultural conceptions about freedom and philosophical debates about the nature of justice are all affected.

That is, the implications of the modern rise of automation go well beyond the question of what happens, in purely material terms, to those people who lose their jobs as a result; it raises questions about our ideas of work, leisure, the purpose of life and the nature of the human, it places our system of values under strain. These are to original observations; my suggestion is simply that we might learn something from studying how a different culture was shaped by similar circumstances. I’m thinking of re-reading Fitzgerald’s Slavery in the Roman Literary Imagination as a starting-point.

There are other possibilities too. Probably trite to consider the issue of slave revolts, though the Romans’ relative success in limiting rebellion largely to small-scale, individual acts of resistance (non-cooperation, dumb insolence, ordering unwanted items from Amazon etc.) is noted. The issue of manumission – why the Romans granted such autonomy to ‘things’, so frequently, and how they negotiated their transition to personhood – is certainly worth thinking about. Still more, the fate of the redundant masses, excluded from meaningful activity and reduced to aspiring to own robots of their own – echoes here of one of Peter Frase’s Four Futures.

There’s a substantial tradition, especially in the nineteenth century, of contrasting ancient slave society with modern capitalism. I always recall the Aristotle quote with which I started from Marx’s evocation of it in Das Kapital: foolish Greek, thinking that machinery would lead to a life of leisure, rather than being the surest method of lengthening the working day! Likewise, “the Roman slave was bound with chains… the modern wage-labourer is bound to his owner by invisible threads”. Manifestly, Marx failed to imagine that the remorseless logic of capitalism might lead workers to be displaced rather than exploited, and that we might be better off thinking of analogies between Juvenal’s “bread and circuses” snark and the joys of social media…

On the one hand, there have started into life industrial and scientific forces, which no epoch of the former human history had ever suspected.  On the other hand, there exist symptoms of decay, far surpassing the horrors recorded of the latter times of the Roman Empire.  In our days, everything seems pregnant with its contrary.  Machinery, gifted with the wonderful power of shortening and fructifying human labour, we behold starving and overworking it.  The new-fangled sources of wealth, by some strange weird spell, are turned into sources of want.  The victories of art seem bought by the loss of character.  At the same pace that mankind masters nature, man seems to become enslaved to other men or to his own infamy.  Even the pure light of science seems unable to shine but on the dark background of ignorance.  All our invention and progress seem to result in endowing material forces with intellectual life, and in stultifying human life into a material force. (Marx, Speech at th3 anniversary of the People’s Paper)