General Average, Human Jettison, and the Status of Slaves in Early Modern Europe

Abstract

This article proposes a transition in Western European thinking on slavery by examining the legality of slave jettison and its indemnification in the seventeenth-century Christian Mediterranean and comparing this with the late eighteenth-century Atlantic. Under the law of general average (GA), a shipmaster may legally sacrifice cargo or parts of a vessel to save a maritime venture from peril. GA then mandates that the costs of this sacrifice be shared proportionally between all interested parties. However, the status of human cargo with respect to pre-modern GA remains unclear, beyond the well-known example of the eighteenth-century British slave ship, the Zong. A jettison, a moment of crisis, forces the slave's dual conception as person and property to be definitively resolved. This article uses historical GA records and early modern jurisprudence on human jettison to shed light on the legal conceptualization of the slave in the two contexts. It finds that seventeenth-century jurisprudence generally ruled against slave jettison and that such a jettison could not be indemnified. In some Mediterranean operational contexts, slaves were excluded from GA altogether. To a certain extent, this finding justifies the conceptual divide historians have placed between Atlantic bondage and earlier forms of slavery.

He asks whether, if something should be thrown overboard at sea, it should be an expensive horse or a cheap little slave. Here our estate inclines us in one direction, our humanity in another…¹

Cicero never gave his final answer to Hecaton's conundrum, posed in Book III of De officiis (On duties). Since the time of the Roman empire (and possibly earlier), it has been considered legal for a shipmaster to sacrifice cargo or parts of the ship during a crisis in order to save the whole venture. The classic example is an act of jettison, in which cargo is cast overboard to lighten the ship during a storm. A legal procedure has also existed for sharing the costs of such a sacrifice fairly between all interested players, with each party contributing to the loss in proportion to their original investment in the ship and/or cargo. Now known as general average (GA), this procedure continued to be used after the fall of the Roman empire in the Christian and Islamic worlds, albeit with significant regional variation, and it remains a fixture of international maritime law today. The animating principle of modern GA is ‘that which has been sacrificed for the benefit of all, should be made good by the contribution of all’.² As Cicero notes, however, the issue of sacrifice and compensation was far less straightforward when the ‘cargo’ in question was comprised of enslaved persons. Could this ‘property’ ever be sacrificed? And should slave owners be obliged to contribute to GAs in turn? A moment of crisis forces the status of the slave to be clarified.

This article uses questions about slave jettison and GA as a means of detecting broader legal and conceptual transformations affecting the category of ‘slave’ across the early modern period. The evidence it presents is largely drawn from the seventeenth-century Christian Mediterranean, and the case-study it examines unfolded in Medicean Tuscany. Yet its findings are also relevant to our understanding of Atlantic slavery. The latter is characterized paradoxically as both entirely novel in its brutality and commodifying intent yet at the same time paradigmatic, with neither supposition being supported through comparative analysis with other forms of bondage. Though these initial findings are necessarily provisional, this article apprehends an important shift in attitudes that should be explored in future research: by comparing seventeenth-century jurisprudence and practice from the Mediterranean with later examples from the Atlantic, it suggests a turning point in European thinking about slavery, seemingly motivated by both racial theories and new notions of property. It finds that most learned normative authorities in the sixteenth and seventeenth centuries seem to have rejected the legality of slave jettison, but that this, in turn, created confusion as to whether slave owners should contribute to GAs at all; on this latter point, there is a dissonance between the learned tradition and some local customary laws. The case-study also uncovers a subtle distinction between the slave as ‘property’ and the slave as ‘merchandise’ which deserves to be recognized and given further attention by scholars. In seventeenth-century Tuscany, the slave was considered the former but not the latter, a distinction that fell away in the later period. The findings thus seem to confirm the existence of a conceptual divide between Atlantic slavery and earlier forms of bondage.

The first section considers the ‘commodification paradigm’ and its effect on the historiography of Atlantic and Mediterranean slavery. The second section presents a case-study from seventeenth-century Tuscany, which, though not a slave jettison itself, turned on the question of whether a slave jettison could be carried out and indemnified. The third will consider learned jurisprudence on the question of slaves and GA from before the eighteenth century. The fourth section will consider the relationship between these learned and customary traditions in the seventeenth century and will argue that a distinction existed in certain cases whereby slaves were considered ‘property’ but not ‘merchandise’.

I

This study of slave jettison and its indemnification speaks directly to the question of ‘commodification’. As Nicholas T. Rinehart has noted, the ‘commodification paradigm’ has become an axiomatically accepted truism in discussions of the Atlantic slave trade.³ Many historians of Atlantic slavery have suggested that, of the many brutalities it perpetuated, it was this ‘objectification’ or ‘dehumanization’ of Africans that was the trade's essential and most damaging characteristic.⁴ Historians have also argued that this commodification – turning a human being into a ‘thing’ – was unique to this particular historical moment, thus distinguishing Atlantic slavery from all previous forms of bondage.⁵ Ian Baucom has posited that Atlantic slavery represented the final stage in the development of modern finance capitalism, now capable of transforming anything, even human beings, into its monetary equivalent.⁶ Other historians have chosen to emphasize how slaves resisted this classification as ‘thing’ by reclaiming human agency.⁷ Yet this commodification paradigm, as Rinehart has shown, is rarely used to an analytical end, but is too often glibly invoked as a form of virtue signalling by modern-day historians; in its current form, it adds little to our understanding of slavery as a historical phenomenon.⁸ Walter Johnson has likewise encouraged historians to move away from repeatedly exposing latent contradictions in the philosophy of slavery.⁹

The Atlantic commodification paradigm has sent distorting ripples towards the historiography of Mediterranean slavery too. Claims to its historical singularity notwithstanding, Atlantic slavery has become the exemplar against which other forms of slavery are measured, often implicitly. Consequently, some historians have insisted on the use of the term ‘captivity’ rather than ‘slavery’ in the Mediterranean case. It was Michel Fontenay who originally argued for this precise deployment of terms, reacting primarily against scholarship that talked misleadingly of ‘white slavery’ in North Africa. When making this argument, however, Fontenay explicitly compares the Mediterranean experience to the Atlantic. Citing the widespread practice of ransoming captured persons in the Mediterranean, Fontenay argues that while a slave in the Atlantic had a ‘value of usage’ and their price was the ‘price of the man’, the Mediterranean slave ‘value of exchange’ and their ‘price’ was the ‘price of liberty’.¹⁰ Fontenay contrasts the transient Mediterranean experience with the ‘dehumanizing’ slavery invented by modernity (i.e. Atlantic slavery) in which the condition of the slave was all too permanent.¹¹ Furthermore, according to Fontenay, while trade in Atlantic slaves occurred between two agents of the same cultural-religious group, ransoming occurred between agents pertaining to two different groups (i.e. Christians and Muslims), thus introducing a ‘political, ideological, and religious dimension’.¹² For Fontenay, ‘this was a form of slavery no doubt archaic, but ultimately quite humane’.¹³ Wolfgang Kaiser and Guillaume Calafat have developed this claim, arguing that ‘captive’ not only avoids improper analogy with the Atlantic case but better reflects contemporaries’ understanding of their own situation.¹⁴

In short, Mediterranean historians too have acritically fallen into the Atlantic commodification paradigm on occasion, albeit from the other side of the divide.¹⁵ First and foremost, Fontenay's argument hardly does justice to the variegated experience of Mediterranean bondage. Not all slaves would be ransomed – one thinks of slaves in domestic service across Italy, for example, or the thousands of slaves who manned the oars of the war galleys – and the idea that this was a ‘quite humane’ process is also questionable.¹⁶ The suggestion that the Atlantic slave trade was monocultural is also flawed, ignoring the role played by West Africans in the procuring and selling of slaves. Above all, however, it departs from the problematic premise that all discussions of slavery in world history need to defer to the Atlantic experience, whilst uncritically accepting that full ‘commodification’ was its defining feature.

This article aims to provide a new and meaningful perspective on this question. Rather than simply asserting that slaves in the Atlantic were ‘commodified’, it uses a comparative framework to identify a broad turning point in European thinking about the slave as ‘commodity’ whilst also acknowledging the presence of ambivalence and local variation. Igor Kopytoff argues convincingly that slavery should not be viewed as a binary ontology (person or thing) but rather as a process: the slave was always a ‘potential commodity’ but underwent de- and re-individualization at various points, with the moment of sale being the only one in which the slave was fully commodified.¹⁷ Furthermore, as Walter Johnson writes, the ‘daily process of the slave trade’ involved ‘many slave trades; many versions of what was happening’.¹⁸ With that said, these productive observations do not answer the question of whether the Atlantic slave trade represented something new in terms of the slave's commodification, a claim that is, apart from anything, central to those theories that see Atlantic slavery as integral to the birth of modern finance capitalism. This is where the legal question of slave jettison and indemnification – whether slaves could be ‘sacrificed’ like goods and whether that ‘sacrifice’ could be legally reimbursed – can offer us new insight. While recognizing that slavery and commodification were processes, the crisis that precipitated a jettison at sea forced the slave's dual identity as person and thing to be resolved definitively. This article suggests that in this respect Atlantic slavery did in fact represent something new. Yet it also demonstrates that the question provoked diverse and often ambivalent responses, and it is not possible to say that European jurisprudence adopted a consistent position on the issue in either period.

The legality or otherwise of slave jettison in early modern Europe remains a vexed question, with scholarship directly addressing the issue fragmented and contradictory. This perhaps stems from the fact that scholars have concentrated on evidence from the 1770s and 1780s, a fissile moment for the juridical status of the slave as the Atlantic trade came under new scrutiny. Based on a survey of late eighteenth-century jurisprudence, Maura Fortunati claims that slaves were not jettisonable.¹⁹ Alexandra Philip-Stéphan, on the other hand, claims that the Lex Rhodia (the section of Justinian's Digest that deals with jettison and GA) considered slaves to be ‘merchandise’, allowing both their jettison and the recouping of their value via GA.²⁰ Meanwhile, the case of the Zong has been particularly influential, despite the fact that a verdict was never reached.²¹ In 1781, the crew of the now-infamous British slaving ship jettisoned 132 sick Africans into the Atlantic Ocean. Falsely stating that water supplies had been running perilously low, the owners of the vessel then tried to claim for this loss from their insurers through a declaration of GA. Scholars have drawn quite general conclusions on the basis of Lord Mansfield's instructions to the jurors to consider the slaves as no different from horses. But when the veracity of the plaintiff's account was called into question the parties decided to settle out of court, saving Mansfield from making a final decision. Some contemporary scholars, moreover, have questioned the legal basis of the likely outcome.²²

The fact that scholars have examined the problem at a moment of particular flux and growing abolitionist pressure is partly responsible for these disagreements. Fortunati's view is principally based on her reading of the late eighteenth-century Sardinian lawyer Domenico Alberto Azuni, who in turn based his arguments on the work of Balthazard-Marie Émérigon. Yet Émérigon's view was highly influenced by Enlightenment thinking about natural law, as well as his own personal aversion to slavery, and he found no support for his view in the work of earlier authorities: the authors he cites in the relevant passage, Cicero and Samuel Pufendorf, come down neither for nor against the proposition.²³

With that said, it is fairly clear from emerging evidence on the insurance of slave cargoes that mercantile practice in the Atlantic often allowed for the indemnification of slaves sacrificed through GA, whatever its technical legality. Since GA contributions were usually covered by insurers unless the contract explicitly stated otherwise, insurance policies can give some clue as to contemporary practice. James Oldham finds that many contained a ‘free from average’ clause for cases of slaves killed during insurrections; this clause excused underwriters from paying GA contributions under a certain threshold, usually between 3 and 10 per cent of the total insured amount.²⁴ As Oldham remarks, small losses on this account were clearly so frequent that underwriters wished to avoid settling frequent small claims. Similarly, Mallory Hope has found some French insurance contracts from the late eighteenth century that explicitly mention the jettison of slaves among the underwriter's liabilities.²⁵ Though no seventeenth- or eighteenth-century documentation of an actual slave jettison has yet come to light (other than the Zong), this is not necessarily surprising: since GA cases were usually handled privately in England, records are hard to come by, while no one has yet studied the GA cases contained in the voluminous records of the French admiralty courts.²⁶ Based on the extant evidence, it does not seem that Mansfield's likely decision went contrary to contemporary practice: for the purposes of GA, slaves were treated like any other cargo.

II

In seventeenth-century Tuscany, the consensus was somewhat different. A case-study demonstrates that there was confusion about whether slave owners should have to contribute to the indemnification of GA sacrifices. This in turn was seemingly based on the assumption that slaves could not be jettisoned and that any such jettison would not be indemnifiable; as will be shown, the two questions were inter-related.

The case itself unfolded in the Tuscan port of Livorno. Founded almost from scratch by the Medici grand dukes in the sixteenth century, Livorno grew to be one of the most significant ports in the Mediterranean in just a few generations. It is best known as one of the first ‘free ports’, where merchants of all religions and nationalities could settle and trade on an equal footing, at least in theory.²⁷ Yet the ‘free port’ was no stranger to human bondage; on the contrary, it was at the very centre of human trafficking in the western Mediterranean. Slavery in the Mediterranean was justified largely on the basis of religious rather than racial difference; many of the slaves brought to Livorno would have been Muslims hailing from the Levant, with some originating from North Africa and a few from sub-Saharan Africa.²⁸ The sixteenth century saw a considerable shift in the patterns of enslavement in Italy: the onset of the corso – the continual, low-level naval conflict between Christian and Muslim ‘corsairs’ that continued after the peace between the Spanish and Ottoman empires in 1581 – both cut off older slaving routes from the eastern Mediterranean and the Black Sea and provided a new source of potential slaves in the form of captured mariners and travellers: male slaves of Levantine origin came to represent the majority of slaves present in Tuscany.²⁹ Some of the slaves would be destined for public service on the galleys, in the arsenale, or in the state biscuit factory; others were purchased by private individuals. Women and girls were purchased as domestic slaves, though their numbers declined from the sixteenth century onwards.³⁰ Across the seventeenth century, around 5 to 10 per cent of Livorno's population was made up of slaves.³¹ Well into the eighteenth century, Livorno remained a crucial market for French and Spanish agents in constant competition to fill the benches of their sovereigns’ galleys.³² The importance of slavery, both actual and symbolic, is attested by the Monument of the Four Moors, constructed in 1626 and placed in a prominent position near the waterfront: Grand Duke Ferdinando I overlooks the port, standing above four chained slaves of various physiognomies.³³

The case in question is a GA claim made in 1671 by Paolo Antonio Pieri, the master of the ship the Madonna di Monte Nero.³⁴ Records of the case are preserved in the archive of the consoli del mare di Pisa (the consuls of the sea in Pisa), who retained significant maritime jurisdiction in Tuscany, including over GA, even after the rise of nearby Livorno ended Pisa's role as a port city. The Monte Nero had come from Izmir, with scheduled stops at Messina, Livorno, and Genoa. Soon after leaving Izmir, it had called at two Aegean islands where it had loaded slaves: twenty-four loaded on the island of Siros by ‘Captain Giovanni Maria Cardi’ and ‘Captain Paolo Barbieri’, and thirty more by ‘Captain Giuseppe Corbara’ on the island of Psara. No further details about these enslaved persons can be gleaned from the sources.³⁵ Though nominally under the control of the Ottoman empire, the Aegean islands in this period were notorious bases for corsairs of both faiths.³⁶ It seems likely that these captains had been patrolling the Levant and were dropping off their captured booty with a merchantman in order to continue raiding in the area. The fifty-four slaves were to be delivered to one Giovanni Francesco Cardi, along with ten bales of coffee.³⁷

Somewhere between these islands and Zante, the crew of the Monte Nero jettisoned a large amount of cargo (though no slaves) to make their escape from a ship that they had thought was a corsair.³⁸ In theory, any jettison had to be made by the shipmaster in consultation with both the crew and any merchants who happened to be present. In reality, merchants were very rarely on board, and storms and other crises did not at any rate constitute an ideal context for considered discussion. Jettisons were almost certainly made by the executive decision of the master; by the late seventeenth century, even official GA accident reports seem to have been abandoning any pretence on this front.³⁹ The Monte Nero then stopped in Zante to make an accident report in front of the local Venetian governor and his counsellors, before carrying out the GA procedure at Messina – the first scheduled port of call. By the time the ship reached Messina, eleven of the slaves were already dead.⁴⁰ It was generally expected that a shipmaster would make his accident report in the first (Christian) port after the incident, hence the stop at Zante.⁴¹ Yet it was at Messina, the first scheduled stop in which merchant receivers were present, where important decisions regarding the framing of the case were made. It was there that the master's account of the jettison was judged to be truthful, and it was the Messinian consuls who included the value of the slaves in the GA calculation when it was worked out how much each interested party would contribute.

With the GA awarded by the Messinian authorities, and with contributions collected from the receiving merchants there, the ship continued its journey to Livorno. While unloading in Livorno it was discovered that a few jettisoned items had been missed from the original Messina calculation: two bales of cowhide and four bales of wool.⁴² The master, Paolo Antonio Pieri, went to the court of the consoli del mare and requested that the cost of these missing items be added into the GA, a request which required a new judgement and calculation. This was somewhat unusual: once a GA had been awarded by a competent local authority it was not usually necessary to involve the authorities of another port. A master could simply present the judgement to any receiving merchants and rely upon a general expectation that the decision would be honoured, even if it had been made in another jurisdiction.⁴³ A transnational procedure could not work without such acceptance. However, once Pieri had discovered that even a few jettisoned items had been missed from the original calculation, he was compelled to seek another GA judgement and calculation to avoid being held liable for their loss. A few months later, this decision was the subject of a petition presented to the Tuscan Grand Duke Cosimo III. The ‘exorbitant Average’ granted by the consoli had caused ‘grave damage’ to the petitioner.⁴⁴ It was perhaps little wonder that the case attracted controversy: the jettison had been worth an enormous 22,745 pieces of eight, with the GA payment amounting to over 20 per cent of each individual's investment.⁴⁵ The petitioner was Giovanni Francesco Cardi, the receiver of the slaves.

The value of the fifty-four original slaves had been estimated for the purposes of the GA at 5,273 pieces of eight, 5 soldi and 8 denari. Cardi's total interest in the voyage was almost 6,000 pieces, and his required contribution stood at 1,237 pieces, 9 soldi and 4 denari.⁴⁶ His complaint was manifold. He made several general criticisms about the way in which the GA had been conducted: that the jettison seemed far too large to be plausible, that the GA ought to have been adjusted in Pisa rather than in Messina, and that he had not been cited at the time of the Pisan procedure.⁴⁷ His main complaint, however, centred on his cargo of slaves. Cardi argued that the slaves should never have been entered into the GA in the first place. This was the first point made in his list of objections:

It has never at any time been the custom that slaves should be placed into Average: on the contrary, doing so would be a manifest and notorious injustice. Furthermore, since slaves cannot be considered anything other than passengers, and for this reason, as passengers, they do not go into GA, slaves ought not to be placed in the calculation.⁴⁸

Cardi then went on to argue that he should not have to contribute for fifty-four slaves when he had only received forty-three; he suggested that the eleven who had died in transit should either not be counted, or that the shipmaster should be made liable for their contribution.⁴⁹ It was made clear, however, that this objection should not be taken to undermine his principal claim that slave owners ought not to contribute at all. In underlining this, Cardi's defence makes a particularly interesting statement, further clarifying that slaves ought not to be considered ordinary merchandise:

If the slaves should be considered as merchandise (which fact is rejected), it is certain that, with eleven of the fifty-four dead…the price of those who have died should also go into GA, or that the master should be condemned to the delivery of those slaves or their equivalent value.⁵⁰

Cardi was joined in his suit by other receiving merchants who submitted their own objections.⁵¹ Here, the apparently long-established practice of slaves not entering into GA was once again foregrounded:

But what removes every difficulty is…the fact that [the consoli] place in the calculation, in respect to Master Cardi, 54 slaves, when it is a well-known fact that slaves are never put into GA, as diverse merchants attest that have received slaves on other occasions, and that when GAs have been carried out the slaves have never been put in the calculation.⁵²

Cardi's assertions and the support they commanded are somewhat surprising because they completely contradict the normative sources on the contribution of slave owners to GAs. Chief among these is the Lex Rhodia de Iactu (The Rhodian Law of Jettison), the provisions concerning jettison and GA contained in Chapter 14 of Justinian's Digest, one of the foundational texts of the ius commune. The Lex strongly suggests that a slave owner is expected to share the costs of any sacrifices made to save the ship. The relevant passage poses an instance of a vessel carrying ‘diverse cargoes…in addition to many passengers, both slave and free’ which performed a jettison. The question is then asked ‘whether the people whose goods, such as jewels and pearls, added no weight to the ship had to contribute like everyone else…whether anything was due in respect of the free passengers, and by what action the matter should be proceeded with’.⁵³ It was decided that ‘all those who had benefited by the jettison must make their contribution…freemen not being valued’, thus implicitly mandating contribution for the value of the slaves.⁵⁴

This stipulation was likewise echoed by seventeenth-century jurists. The Genoese maritime jurist Carlo Targa explicitly writes in his 1692 work on maritime law that ‘gems and other precious things are also valued, and cash and the value of slaves (though only those who are infidels) are placed [into contribution]. For while these things are not subordinate to the jettison, not providing relief, they nevertheless enjoy its benefit.’⁵⁵

Giuseppe Casaregi, another influential jurist whose career encompassed the Genoese, Sienese, and Florentine Ruote, writes similarly in Discursus legales de commercio, first published in 1707:

And into contribution comes all merchandise and all things in the ship, and everything that was saved. Even light or very light merchandise and possessions, though they weigh down the ship little or not at all, contribute the same: Slaves, of course, gold coin, money, gems, pearls, rings, precious stones, and everything of this sort.⁵⁶

The same provision even made its way into the Genoese statutes of 1589, which called for the contribution of ‘all things existing in the said ship at the time of the jettison, including money, gold…slaves, male and female, horses, and all other animals’.⁵⁷

While the works of Targa and Casaregi slightly postdate the case in question, the Roman Lex Rhodia de Iactu certainly did not. And yet the Lex Rhodia was never referenced by Cardi's opponents in the case, nor by the two consoli del mare when they were called on to justify their original decision. The closest that the consoli came to making the argument outlined in the Lex Rhodia was their remark to the effect that ‘slaves, as they receive a valuation, in contrast to free people, ought to be placed in the Average as merchandise’.⁵⁸ But the consoli made no reference to the Lex Rhodia nor to the fact that Cardi had benefited from the sacrifice. Instead, they argued by analogy, stating that ‘since slaves pay import taxes in the same way as merchandise, it does not stand to reason that it should be different in cases of Average’.⁵⁹ Paolo Antonio Pieri, the shipmaster who had claimed the original GA, likewise made no effort to counter Cardi's central claim. His only comment on the slavery question was to remark that ‘it does not matter that Master Cardi had the Average upon fifty-four slaves when in Livorno arrived no more than forty…on this point he has no quarrel with the consoli in Pisa, but rather with those of Messina (even though these have done no wrong)’.⁶⁰

Rather than rebutting Cardi's point in principle, the master simply passed the buck from the Pisan consoli to those of Messina, over whom no jurisdictional power could be exercised. The consoli likewise argued that the case could not be reopened because of the practical difficulties that would ensue: since the original GA had been carried out in Messina, and since the receivers in Messina had already paid their contributions, it was simply not feasible to reverse this process.⁶¹

Historians must be careful in taking assertions of ‘custom’ at face value: as Guido Rossi has pointed out, mercantile expressions of custom were open to abuse, and an influential merchant might easily martial enough support from the community to lend an air of ‘customary’ authority to an entirely novel position.⁶² In this case, however, there are several grounds for giving credence to Cardi's remarks. First, there is the eventual decision made by the Florentine Ruota, the highest civil court in Tuscany, to whom the petition was delegated by the grand duke.⁶³ Though we cannot be sure on what basis the judges of the Ruota deliberated, they nevertheless ordered on 8 February 1672 that the case should be reviewed.⁶⁴ ‘Review’ here was euphemistic, a concession to the fact that judgements of the consoli were technically final. Secondly, we have the fact that the existence of the custom went uncontested by the opposing side. Finally, a preliminary search of the Pisan archive has revealed no other seventeenth-century GA case in which slaves were included in the calculation, even though Livorno possessed the most important slave market in the Christian Mediterranean after Malta.⁶⁵ The only reason that the slaves came to contribute in this case was that the GA was originally carried out at Messina, where such prohibitions presumably did not apply; when the GA was readjusted in Livorno, the slaves were not removed, but the Pisan consoli had not actively chosen to include them. It is thus highly likely that slaves were in fact excluded from GAs as a general rule.

III

Why did the Tuscan merchants consider it proper to exclude slaves from contribution? Digging deeper into Cardi's claim casts interesting light on the status of slaves in seventeenth-century Tuscany. This section will show how Livornese GA customs maintained strict notions of reciprocity: property that could not be indemnified through GA, or which could be sacrificed at all, could not, in turn, be entered into GA as a contributing object. We will see how sixteenth- and seventeenth-century jurisprudence generally held that slaves could not be jettisoned. The operational conditions in the Mediterranean, meanwhile, meant that the practical need for slave jettison was far less acute than in the Atlantic.

Logical justification of the custom of excluding slaves is never explicitly made in our archival documentation. This is not necessarily surprising. After all, the petitioners did not have to justify custom, they simply had to show it existed. Ultimately, there seem to be two rationales at play here, which are substantially inter-related. The first is that the slaves were not considered to be ‘merchandise’, a thing that cast doubt on their eligibility for GA. The second is that merchants seem to have internalized the idea that GA was strictly reciprocal, and that things that could not be sacrificed and indemnified via GA could likewise not be considered eligible for contribution to GA. In other words, if slaves could not be jettisoned – or if their owners could not receive the protection of indemnification after a slave jettison – those same owners should not be compelled to contribute on account of their slaves either.

In both modern and Roman GA, the obligation generated between the one who has lost property and the one who has benefited is a one-time obligation on account of the sacrifice made. Hence, the Roman law held that slave owners had to contribute because they had benefited financially thanks to the sacrifice. Early modern Livornese merchants, on the other hand, saw GA more as a form of reciprocal mutual insurance, a kind of compact between players in the maritime-business community. This is well illustrated by a GA case that unfolded thirty years after the dispute over Cardi's slaves. This concerned a GA requested by Antonio Fogazza, the master of the Sante Anime del Purgatorio, on his ship's return to Livorno from Tétuoan in 1700. The damages to be repartitioned included the costs of a jettison and expenses incurred during a period of detention in Gibraltar by two French warships.⁶⁶ The GA was disputed by the two merchants, Iuda Crespino and Ioseph Molco. The parties could not decide whether cargo owned by the master, which had supposedly been stored above deck, should contribute to the GA. The master claimed that his goods should not contribute, because ‘these were loaded under the bridge and not in the hold, for which fact, according to Chapter 183 of the Llibre del consolat de mar, they are excluded’.⁶⁷ The Llibre del consolat de mar was the most important collection of written maritime customs in the western Mediterranean and was frequently cited in cases coming before the consoli.⁶⁸ It had been compiled in Catalonia in the fourteenth and fifteenth centuries, bringing together various rules and practices that had been employed in the western Mediterranean from the twelfth century onwards.⁶⁹ When we turn to the Llibre, however, we do not find it stated that cargo stored above deck does not contribute to GA. Rather, it says that if cargo is loaded above the hold without the permission of the merchants, then it is not indemnifiable and that the master will have to bear the whole cost of any loss or damage.⁷⁰ In effect, it suggests that if the master's cargo stored above deck had been jettisoned, he could not have benefited from GA, because he had assumed sole liability. And yet the master believed that the fact that his cargo was stored above deck excluded it from contribution as well: in short, that exclusions applied to sacrificed cargo also applied to contributing cargo.

This logic was accepted by the merchants, who did not initially object to the principle of the exclusion. Their problem was simply with the veracity of the master's account. The merchants claimed that such a quantity of merchandise would not have fitted on the deck and that the master was thus lying about where he had stored his goods.⁷¹ The master, in turn, was forced to produce an affidavit signed by the customs officials to the effect that the goods really had been stored above deck.⁷² It was only after both sides had submitted several lists of objections and counter-objections that the merchants suddenly changed tack. In their third and final list of objections, they instead argued, perhaps after seeking out new legal opinion, that things stored above deck should contribute after all because ‘by means of the jettison they had been saved’.⁷³ In other words, they switched from the logic of strict reciprocity they had previously accepted – that cargo loaded on deck is excluded from GA in any respect – to the logic of the Lex Rhodia: even though things loaded on deck could not have been indemnified through GA, they nevertheless should contribute because they benefited from the jettison. In their final judgement, the consoli sided with the merchants, but it is clear that the merchants themselves were not familiar with the logic of the Lex Rhodia until very late on in the process.⁷⁴

The fact that both parties were initially convinced that the Llibre del consolat de mar excluded cargo stored on deck from both sides of the GA equation suggests that seventeenth-century Livornese merchants were not viewing GA through the prism of sacrifice and benefit but rather through the lens of a do ut des contract. GA in this conception was closer to a pact of mutual insurance than to the idea of voluntary sacrifice contained in the Lex Rhodia. While GA was not strictly contractual, it represented a kind of social compact among the maritime community: I contribute to your damages today in the knowledge that, on another occasion, the situation will be reversed; if, however, I receive no protection, it is not right that I should be called upon to protect you in turn.

If, in mercantile thinking, GA depended on strict reciprocity then the contribution of slaves to GA was linked to their ‘jettisonability’. If slaves could not be jettisoned, or if their value could not be recouped after a slave jettison, slave owners should not have to contribute to GA either since their property did not receive the protection of mutual contribution. In the Mediterranean, moreover, it was highly unlikely that circumstances precipitating a human jettison should occur. Recorded human jettisons were never carried out to lighten the ship but were occasioned by a lack of supplies, disease, or by slave rebellions, in which slaves were precipitated overboard in the struggle or were deliberately drowned as a punishment and example to others.⁷⁵ Slave jettisons were thus ultimately the result of the difficulties of oceanic navigation, the desperate situation of the enslaved, and the industrial scale of the operation, pressures which were far more prevalent in the Atlantic context. Though there are recorded instances of slave rebellions in the Mediterranean, proximity to land, the much shorter journey times, and the slaves’ knowledge that ransom was a possibility in some cases meant that these were not pressures that affected vessels carrying Mediterranean slaves to the same degree.⁷⁶ We can say with some confidence, therefore, that slave jettison must have been an exceptionally rare occurrence in the Mediterranean.

Whether slaves were considered jettisonable in theory is another matter. Though discussion on this point was sometimes characterized by a degree of ambivalence, most normative sources of maritime law in the seventeenth century either forbade the jettison of slaves or, at the very least, stipulated that slave jettison should not result in any compensation for slave owners. It was this latter position that was taken by the Lex Rhodia. Though ius commune jurisprudence did not necessarily prevail over customary law and merchant usage in early modern maritime disputes, the Lex Rhodia nevertheless remained a most important reference point for learned jurists, including English lawyers writing on maritime law.⁷⁷ Passage 14.2.2.5 of the Lex Rhodia stipulates that ‘no valuation is to be put on slaves who have perished in the sea, any more than if they had sickened and died on board, or thrown themselves in’.⁷⁸ This apparently straightforward remark has been variously interpreted by modern scholars. Some have assumed that ‘slaves who have perished in the sea’ refers only to slaves who die through accidental drowning. Hassan Kalilieh, like Alexandra Philip-Stéphan, maintains that the Lex Rhodia allowed both the jettison of slaves and the recouping of their value via GA: he thus takes this passage to refer to slaves who have abandoned ship.⁷⁹ Émérigon interpreted it as referring to slaves who have perished as the result of ‘perils of the sea’ – as he claimed that slave jettison had never been considered legal, he could hardly read it otherwise.⁸⁰ Other scholars, such as Olivia Remie Constable, have interpreted the ban on valuation as referring to deliberately drowned – that is, jettisoned – slaves.⁸¹

In the absence of any qualifications on the legislator's part, the most logical approach is surely to take the Lex at its word: the valuation of slaves who have been drowned, either accidentally or on purpose, is completely ruled out. This makes it impossible, under the Lex Rhodia, to receive compensation for jettisoned slaves through contribution, since this would necessarily involve valuing them. In this respect, we might consider how the reference to slave suicide is expressed in the original Latin: aut aliqui sese praecipitaverint. Literally, this translates as ‘or those who have thrown themselves’. The phrase ‘into the sea’ is clearly required to complete the sense. It therefore depends on the in mare from the first half of the sentence. The two phrases are placed in antithesis: one where the slaves have ‘thrown themselves’ in mare, the other in which the slaves have fallen (or been thrown) against their will. This effectively rules out Khalilieh's reading in which the slaves have jumped out of the ship in an attempt to reach land.

The Lex thus suggests, first of all, that the jettison of slaves was not prohibited by Roman law. This supposition is in accordance with the earlier quotation from Cicero. Had slave jettison been legally impermissible, Cicero, a lawyer by profession, would surely have mentioned it. Instead, Cicero frames the dilemma as a purely moral issue. At the same time, however, it seems that slave jettison was discouraged; one would like to think that the refusal to compensate slave owners for a human jettison was designed to discourage such an act, though this may simply be wishful thinking.⁸² What is clear is that GA cannot be used as means of recouping the value of the slaves.

When this passage was transmitted into early modern jurisprudence, most jurists either seem to have rejected the idea of human sacrifice or to have not even considered it a possibility. The latter stance was adopted in Pieter Peck the Elder's commentary on the Lex Rhodia, produced in 1556.⁸³ (This work was in fact cited by Cardi's merchant allies in the case of the Madonna di Monte Nero, though not on the subject of slaves.)⁸⁴ Peck offers a rather unexpected interpretation of passage 14.2.2.5 – one which we should probably dismiss as a guide to the Lex's original intentions, but which is nevertheless interesting in understanding the contemporary view on slave sacrifice. Peck writes that,

when a slave has perished in the sea, ejected by means of ropes when the mountains of water churn (as he [Ovid] says), it should be recognized that this is carried out on account of his going about his duty to the ship, rather than that he was ejected on account of the common benefit.⁸⁵

As Peck elaborates, it becomes clear that he considers the slaves of 14.2.2.5 to be those who are working the ship, rather than slaves in transit.⁸⁶ In fact, when commenting on this passage he sometimes uses the word ‘servi’ (the standard word for slaves) but sometimes refers to them as ‘servitores’ (those serving/servants).⁸⁷ Peck then alludes to another passage in the Lex Rhodia in which natural wear and tear to the ship – not eligible for GA – is likened to the damage sustained by a blacksmith's hammer, for which the customer would not be expected to contribute.⁸⁸ This is taken to underline his contention that the drowning of a servant/slave in the course of carrying out his duties cannot be put into GA. He does, however, go on to admonish those masters who set too much sail during a tempest, leading to the ejection of ‘those serving’, presumably here referring to slaves charged with setting the sails in heavy weather.⁸⁹

In the context of the sixteenth-century Low Countries, it appears that the idea of a slave jettison – or of slaves as a cargo more generally – was not even considered. Peck reduced the multiplicity of ways that a slave might perish in the sea to what he considered the one morally acceptable set of circumstances: slaves drowned while carrying out necessary duties during a storm. It is not impossible, of course, that the Lex itself should also have been referring to slaves who were sailing the ship, though this is not likely. Since the Lex mentions those who have ‘sickened and died’ and those who have committed suicide in the same breath as those who have drowned, it seems fairly clear that this is not a reference to slaves who are going about their daily duties but to a context of transit.

Arnold Vinnius made notes on Peck's commentary in an edition of 1647, republished in 1668.⁹⁰ He offers a more detailed appraisal of the passage:

Slaves, who perished in the sea in that time that the ship was labouring, do not thus perish because they are thrown into the sea for the sake of lightening the ship (for men are not thrown into the sea, but rather merchandise is, on account of men needing to be saved).⁹¹

Vinnius, unlike Peck, touches on the possibility of a deliberate jettison of slaves, if only immediately to dismiss it. Here we have our first concrete and indisputable reference to slave sacrifice. The first and most obvious point is that Vinnius claims that slaves are not jettisoned. This is not so much a stipulation as a simple observation. Vinnius does not use the subjunctive here but rather the indicative (projiciuntur instead of projiciantur): it is not that men should not be thrown into the sea; it is simply that they are not. Secondly, it is striking that Vinnius places slaves in the category of ‘men’, in clear opposition to jettisonable ‘merchandise’. Exactly the opposite logic prevailed in cases such as the Zong, where it was decided that the slaves had to be treated as any other cargo.

Vinnius was not the only seventeenth-century jurist to touch on the question of human sacrifice (rather than slave sacrifice, specifically). Reinhold Curicke discussed whether ‘in certain danger of shipwreck or extreme hunger, for the salvation of the greater part of the ship, it is licit to throw someone into the sea’.⁹² Curicke begins from the Book of Jonah, first noting that the sailors who were with Jonah, though they were gentiles, recognized that ‘it was not licit to throw even the most unimportant man into the sea to save the cargo…however precious’.⁹³ Even though Jonah admitted to being the cause of the storm, the sailors did not jettison him until they had no other option. ‘How much less appropriate is it then’, asks Curicke, ‘for Christian men to throw another Christian sailor overboard?’⁹⁴ Likewise, when the apostle Paul was travelling to Rome and his ship encountered a storm, the seamen made three jettisons but did not once consider jettisoning a man. Having established this, however, Curicke makes something of a turn, noting that ‘necessity…is outside of reason’ and that supreme dangers always give pardon of fault.⁹⁵ Quoting the Pharisee Caiaphas in John's Gospel – ‘you do not realize that it is better that one man should die for the people than that the whole nation perish’ – he goes on to conclude that if ‘the greater part of the mariners might be saved, it is licit that they should throw one of their number into the sea, and that this should be used as a solution of the last resort’.⁹⁶ He counsels the use of lots for choosing who should be cast overboard.⁹⁷ Curicke does not explicitly consider the position of slaves. It is unclear where slaves would hypothetically fit into this scheme, and whether they should be included in the casting of lots: since slaves in the Mediterranean were almost by definition not Christian, perhaps Curicke would have sided against them on account of ‘necessity’.

A contrasting and more definite position is offered by the Flemish commentator Mattheus De Vicq in his gloss of Quentyn Weitsen's Tractatus de avariis. The Tractatus was published sometime between 1554 and 1563, with De Vicq adding his commentary in 1617, and was destined to become one of the most important written normative works on GA across Europe. The first known Latin edition dates from 1672, and Giuseppe Casaregi appended Weitsen's work to the 1729 and 1740 editions of his Discursus legales de commercio.⁹⁸ Like Curicke, De Vicq advocates a human jettison in a situation of dire necessity: ‘it is harsh that the drawing of lots should be established between the men in such a case but even harsher that all should perish for the sake of no one man perishing’.⁹⁹ He also remarks, however, that the seamen ‘should not hesitate over humanity, that is, the kinship which nature creates between us, and prefer personal property, even a very precious horse, over a common slave, who is, however, a man’.¹⁰⁰ Like Vinnius then, De Vicq makes no distinction between those men who are slaves and those who are free and infuses the ambivalence of Cicero with a new certainty in favour of the slave.

While we might doubt whether such noble sentiments of common humanity between seaman and slave would have prevailed in practice, these arguments nevertheless demonstrate that the jettison of slaves (other than in their capacity as fellow human beings) was not generally considered acceptable by seventeenth-century jurisprudence. Though Curicke's ambiguous remarks could arguably have been used to justify a slave jettison, he himself did not do so, while De Vicq and Vinnius explicitly argued that slaves should not be treated differently from others on board the ship when it came to human sacrifice. What is most striking about the seventeenth-century jurists’ discussion of human/slave jettison, however, is the terms in which the problem is framed. Most of the remarks presented here assume that any putative human jettison would take place during a storm. Yet human jettisons in real life were rarely, if ever, occasioned by a need to lighten the ship. Human bodies weigh little compared to anchors, hawsers, or sacks of grain. It was, ironically, the slaves’ very human qualities – their susceptibility to illness, their need for provisions, their natural desire for liberty and willingness to resist – that occasioned their jettison. Sixteenth- and seventeenth-century jurists were thus not thinking in terms of real-world examples. Only Curicke mentions the possibility of imminent starvation as another possible reason for a human sacrifice. These deliberations were academic thought-experiments, responding to literary stimuli rather than operational reality; they were either moral musings drawing on biblical examples or attempts to make sense of a seemingly anachronistic passage in the Lex Rhodia. If slave sacrifices were indeed already happening in the context of the growing Atlantic trade, they made no impression on seventeenth-century European legal thought.

IV

The juridical discussion on the problem was thus quite literally academic and demonstrated a fair amount of ignorance of or lack of interest in slavery in the Atlantic. Nevertheless, despite its limitations, it generally assumed that slaves were not jettisoned like other property. A whole range of influential jurists from across Europe either rejected the jettison of slaves (at least in their capacity as property) or did not even consider the question because the thought did not apparently occur to them. Even the Lex Rhodia, though it seems to accept slave jettison in principle, prevented this sacrifice from being indemnified via GA. Slaves could not thus simply be treated like other property for GA, and it is this point that was the animating principle behind the Livornese custom that Cardi and his associates so vigorously defended.

Slaves in seventeenth-century Livorno, moreover, occupied an idiomatic conceptual space: though they were property, they were not always reduced to mere ‘merchandise’. This distinction has not generally been apprehended in modern-day scholarship, where the slaves’ classification under the Roman law of ‘things’ has too often been taken to resolve all outstanding questions.¹⁰¹ Their separateness from the category of ‘merchandise’ is clearly reflected in the language used by all sides during the dispute between Cardi and Pieri. We might here recall the wording of Cardi's first, original objection: ‘slaves cannot be considered anything other than passengers’; and then later and even more explicitly: ‘if the slaves should be considered as merchandise (which fact is rejected)’. At no point in the entire dispute were these specific assertions challenged. The master's counsel did not even consider the point. The consoli did challenge the overall argument, but not by claiming that slaves are merchandise, but rather by arguing that slaves are considered like merchandise for the purposes of GA: ‘slaves, as they receive a valuation…ought to be placed in the Average as merchandise’ and ‘since slaves pay import taxes in the same way as merchandise it does not stand to reason that it should be different in cases of Average’.

This categorization does not contradict the written norms; in fact, we can often find written jurisprudence making the same distinction. We have already seen in the commentaries of De Vicq and Vinnius that a slave was considered a man, and Vinnius goes as far as to place slaves, along with all other men, in antithetical opposition to ‘merchandise’. This is supported in the work of the commercial lawyer Juan de Hevia Bolaño, whose writings we find cited in the case of the Madonna di Monte Nero to support the contention that jettisons must be made via a consultation. Bolaño's volume was first published in Spanish in 1617, and a Latin edition was published in Florence in 1702. He writes that ‘slaves, or servants, are not merchandise, because men, endowed with reason, are not contained within this term’.¹⁰² This refusal to place slaves in the category of ‘merchandise’ is even supported by Roman law itself: ‘Mela says that men are not to be given the label “merchandise”, and on account of this dealers are not to be called merchants but slave-sellers, and rightly so.’¹⁰³

Alexandra Philip-Stéphan has argued that the slave's classification under the law of things (res) was ‘inescapable, fatal’, but this does not appear to have been the case in seventeenth-century Livorno, nor, in fact, in the corpus of Roman law.¹⁰⁴ Slaves’ designation as res is disputed in some parts of the Roman law corpus and in others is laced with caveats. Ulpian dismisses it entirely, claiming explicitly that ‘human beings are one thing, objects another’.¹⁰⁵ The Roman law on usufruct (that is, the entitlement to the yields of leased property) is clear that the offspring of slaves are not to be treated like the offspring of livestock, though it fails to offer a coherent explanation of why this should be so.¹⁰⁶ In his Commentaries, the Roman lawyer Gaius actually contradicts himself in this respect, first explicitly stating that slaves are part of the ‘law of persons’, and then later placing them in the ‘law of things’.¹⁰⁷ The ambiguous status of slaves has been a recurring feature in European jurisprudence. Lord Mansfield's decision to come down in favour of the slavers in the case of the Zong, though it channelled the logic of the contemporary Atlantic trade, was hardly supported by historic jurisprudence on the question of slave sacrifice.

The original rationale behind this property–merchandise distinction is not certain: W. W. Buckland suggests that the use of ‘slave-sellers’ rather than ‘merchants’ stemmed from a recognition of the slave's humanity, though he claims that the distinction was of little practical importance in Roman law.¹⁰⁸ In an early modern context, however, it clearly had an impact on the unfolding of GA cases. Moreover, it reflects a greater willingness to see the slave as a unique and singular type of property that was subject to its own set of rules, a singularity that sprang directly from recognition that the slave was both a human being and a ‘thing’. This was a context in which Walter Johnson's ‘latent contradictions in the philosophy of slavery’ are hardly so evident. Whether we would want to attach a positive moral valence to this state of affairs is doubtful. Commodification might strike us as heinous, but the seventeenth century could quite unproblematically consider slaves to be human beings and property at the same time precisely because this was a society that did not hold all human lives to be equally valuable and that saw inequality not as a regrettable contingency but as part of the natural order.¹⁰⁹

Though we might still wish to temper ontological claims about the ‘commodification’ of slaves in the way Rinehart and Johnson suggest, the Atlantic trade thus did in fact represent something new as far as the status of slaves was concerned: the category of ‘slave’ lost its singular status and was aligned with the logic of ‘merchandise’.¹¹⁰ The case of the Madonna di Monte Nero displays strong contrasts with GA cases from the late eighteenth century. In the case of the Monte Nero, the slave owner, Giovanni Cardi, argued successfully that his slaves were not merchandise in order to avoid contributing towards a GA: his counterparts in the case of the Zong successfully argued the opposite in order to be indemnified after a slave jettison. In the case of the Monte Nero, the slave was never put in the category of merchandise, even by Cardi's adversaries. By the late eighteenth century, the slave's status as merchandise was taken as a given. As Ian Baucom notes, in choosing to dispute the ‘necessity’ of the action, rather than the principle of it, the insurers of the Zong had already conceded the fundamental point: ‘slaves were, as a matter of law, commodities just like any other’.¹¹¹ Or, as one of the lawyers in the case put it: ‘this is the case of Chattels, of Goods…it is the case of throwing over Goods’.¹¹² Neither the permissibility of the sacrifice nor the slaves’ status as merchandise was questioned.

Several caveats should be added. Within the framework of this broad shift in opinion, there was clearly room for variation. The case of the Monte Nero cautions against excessive generalization on the subject of ‘Mediterranean’ slavery.¹¹³ In this particular case, we cannot even generalize about slavery within the Italian Peninsula. While slaves were excluded from GA entirely in Livorno, slave owners appear to have contributed for their slaves in both Genoa and Messina. Nor do broad conceptual differences between the Atlantic and Mediterranean cases necessarily demand the use of the term ‘captive’ in the Mediterranean as Fontenay suggests. To differentiate strongly between the two terms makes concrete a division that was only ever a question of perspective. Roman law does indeed have rules for the servus and rules concerning a captivus, but the distinction rests only in the subject's relationship to the legislating society. The law of the captivus deals with questions arising when the subject – a citizen, from within the legislating society – has been forcibly removed by an alien society, leaving family and property behind; the law of the servus deals with a subject who has been forcibly removed from the alien society and transplanted into the legislating society. The two terms are effectively interchangeable and their use largely depends on the perceived relationship of the bonded person to the speaker's society.¹¹⁴ ‘Captive’ was certainly used in a Mediterranean context, though it is striking in this case that the enslaved persons are referred to only as ‘slaves’ (schiavi).¹¹⁵

The case of the Madonna di Monte Nero raises questions about when these changes occurred, why they occurred, and if they had to do with the Atlantic context, race, the players in the slave trade, or the emergence of a more general process of commodification associated with the industrious revolution and commercialization. Future research should aim to isolate the precise moment at which this shift occurred and to establish the extent to which racial theories and new discourses about property worked separately and in combination to provide its ideological underpinnings. Christian Europeans were seemingly unwilling to sacrifice one group of persons in the seventeenth century (enslaved Muslims, for the most part) but willing to sacrifice another (black Africans) in the eighteenth. It is difficult to imagine that the dehumanizing discourse of scientific racism did not play a part here, but nor should we discount the influence of new notions about private property. As Amy Dru Stanley notes, the new pervasiveness of commodity relations in the late eighteenth century prompted contemporaries to think more urgently about the limits that bounded commodity relations and commodification.¹¹⁶ New dichotomies emerged: market and home, prostitution and marriage, slave and wage labour. The singular, customized position of the seventeenth-century Tuscan slave sits awkwardly in this ontology.

Tuscany and England were, of course, two states with different legal traditions (though less different than is often claimed as far as maritime law is concerned) and very different histories of enslavement.¹¹⁷ Future research might concentrate on states which straddled the Mediterranean and Atlantic divide – Portugal, or even Genoa, a city closely involved in Mediterranean slavery whose merchants also managed many of the monopolistic asientos created for the Spanish Atlantic trade in slaves.¹¹⁸ Furthermore, as more historical evidence of slave insurance in the Atlantic trade continues to be brought into the open, especially from the period before 1750, it may be possible to reconstruct better the ideological and legal transitions that resulted in new, more definite forms of commodification.