1. Universality and lex mercatoria – a short mention

1To state the obvious, few merchants ever felt the need to postpone trading until a series of general normative propositions regulating commerce was already in place. In other words, at its origins commercial law was customary. Trying to investigate the genesis of what is sometimes described as a spontaneous legal order is the prerogative of the legal philosopher and the economic theorist, from Adam Smith to Friedrich von Hayek, but not of the legal historian, for the sheer impossibility to reconstruct a state of affairs on the basis of fascinating conjectures lacking any hard fact. This limit of course does not mean that it may not be violated – only, that it is not productive to do so. Nonetheless, manyscholarssought to explain this original state. To do that, most relied on the so-called lex mercatoria: a set of universal rules produced directly by the merchants for themselves without the intervention of public authorities.¹ Those rules, being wholly endogenous to the mercantile community, were significantly different from those applicable in a law court,² and governed mercantile transactions for centuries, until the modern state began to assert its law and to impose it over other normative sources, especially customary ones.³

2Such theories have been proved time and again to lack any historical foundation. No universal set of rules was ever produced by merchants, whose customs differed sensibly from place to place. Whenever rules on trade emerged, public authorities (a term necessarily wide and articulated in the pre-modern world) were normally involved in their making as much as in their application.⁴ What we find in the sources is a highly variegated scenario: by and large, a series of layers of rules, ranging from general, supra-territorial principles to specific local customs. Their combination sometimes bestowed upon mercantile rules some degree of similarity but never uniformity.⁵ If some general principles are often attested as applied across (parts of) Europe, few specific rules applying in a given market may be found even in neighbouring ones.⁶

3Broad-brush pictures of pre-modern commerce tend to highlight similarities among mercantile customs. Magnifying those similarities and overlooking details, such general accounts portray trade rules as uniform. The next step is an easy one to take: moving from uniformity to universality. Scratching only the surface, those general pictures often fail to recognise (or tend to downplay) the great variety of actual solutions attested locally and regionally, drawing conclusions as general as they are inaccurate.⁷ The most extreme of such solutions portray general principles as immediately applicable as if they were specific rules, or (which is the same) elevate specific rules to the rank of overall principles. This is how a universal mercantile law comes into existence.

4Most literature on the lex mercatoria seeks to provide some historical foundation to modern attempts to deregulate commercial practices.⁸ The focus is, therefore, not on the past but on present-day law. This probably accounts for the rather flexible attitude towards historical sources, as well as for the tendency to cite previous studies as actual evidence of a past that is being imagined.⁹If the result is somewhat confusing, it is however a confusion that builds a critical mass. This vagueness-cum-footnotes is difficult to disprove. As the normative universality claimed by the adherents of the lex mercatoria may be confuted only by punctilious reference to a series of specific local rules, such references are often taken as localised exceptions to the general rule, and so easily dismissed. Perhaps this explains why the lex mercatoria debate resembles a scholarly hydra: for each head one laboriously seeks to chop off, two more will grow back. Even what is probably the most accurate attempt to disprove its historical existence, Emily Kadens’ lengthy and excellent article The Myth of the Customary Law Merchant,¹⁰ was not exempt from criticism.¹¹ The efforts to debunk the idea of a lex mercatoria seem to have only helped to keep alive the debate surrounding it. To some extent, this is unavoidable: proving a negative is impossible by definition.¹² Only by showing what the actual norms underpinning medieval and early-modern trade were could one put an end to this never-ending saga. But, as those very norms are both extremely difficult to assess with precision and remarkably variegated across space and time, complex and detailed studies on specific segments of medieval trade cannot erase the grand narrative of a universal set of rules. Thus the debate goes on, and well-established legal historians, perhaps despairing of ever bringing it to an end, sought at least to save trees by publishing further excellent confutations of the lex mercatoria on the internet only.¹³ As there is little that may be added to the debate, this article will not linger much on the point – if only to save some trees.

5As a scholar of magic law put it, ‘customary law should be created for magicians in a similar way to Lex Mercatoria.’¹⁴ The statement was made without any irony, and it is useful in that it goes straight to the core of the lex mercatoria idea. So, the same scholar goes on, ‘[t]he Lex Magica arose in a specific context similar to the way Lex Mercatoria came to be: a community of professionals created a body of common law because there were no traditional legal rules efficiently regulating the magicians' activities.’¹⁵ This might be the best definition both of lex mercatoria and of the scholarly approach in its favour – the idea that, at some point after the high Middle Ages, merchants in Europe created a universal set of rules to regulate their activity. Whether this creative process took place by way of positive norms (i.e., of rules imposed, the typical top-down approach of legislation) or it consisted in the reiteration of behavioural patterns (i.e., the bottom-up approach normally associated with customs), those rules should have applied across Europe, or at least within large parts of it. Universality lies at the very core of the need for a lex mercatoria: had it been local or regional, it would have contradicted the very reason for postulating its existence – efficiency.¹⁶ A local set of rules different from those of the next large market would have created just as many problems for supra-local exchanges as it would have solved for internal ones. Trading across different markets, merchants would have had little interest in developing sets of rules different and potentially conflicting for each market. Or so the story goes.

6The analogy with the lex magica is more than a simple provocation. Magicians ought to abide by a deontological code, prohibiting among other things the appropriation of the authorship of tricks invented by their colleagues, or divulging them to non-magicians. The enforcement mechanisms are wholly internal to the community of magicians – access to public (i.e., state-backed) conflict-resolution and enforcement mechanisms is dubious at best. Within this community, ostracism seems to be the main threat to the dishonest magician.¹⁷ In the case of the magic community, however, sanctions are weak, given both the lack of internal hierarchical structures within societies for magicians, and that membership is not compulsory for magicians to practice their trade. Specifically, unruly magicians may well appropriate other magicians’ tricks and use them in their shows, or sell them as if they were their own, with little fear of sanctions. Currently, it seems, the only effective sanctions to dishonest magicians are those provided for by intellectual property rights¹⁸ (insofar as magic tricks can be the object of IP rights, which is a problematic issue).¹⁹ The implied argument seems to be that, if magical societies could enforce their own rules, they would be more effective, and this would be beneficial both to the magic community and, ultimately, also to its individual members.

7This line of reasoning is helpful in assessing the fascination of some scholarship with the lex mercatoria idea. Ultimately, its strength lies in the arbitrary conjunction of an axiom and a fact. The axiom is that a set of universal rules regulating trade is beneficial to those engaged in supra-local trade. The fact is that merchants did trade outside their own market before the establishment of a complex set of state-backed rules governing such trade. The best way to justify that fact is, thus, the application of the axiom.

2. Universal principles? The case of general average

8One of the reasons why the concept of a universal lex mercatoria did take root is that, all in all, the main rules underpinning trade often look fairly similar across time and space. It is admittedly difficult to argue that different societies in different places at different times all came up with the same rules by sheer coincidence. It would seem far more likely that the same rule progressively spread along with the traders who used it. It is a matter of common sense. In reaching this conclusion, however, common sense is unduly encouraged by two mechanisms we often use, whether consciously or not. The first is the tendency to blur together specific rules with more general principles.²⁰ This tendency is all the more natural when the sources are not as clear and precise as one might wish. Medieval mercantile sources, as anyone familiar with them knows all too well, are a case in point. The second mechanism is to associate ‘comparison’ with ‘equation’ – two concepts clearly distinct in theoretical dissertations on comparative legal methodology, but far more complex to disentangle in practice.²¹ Only by avoiding the easy shortcut of equating what looks similar can one distinguish between actual influences leading to the same or very similar rules on the one hand, and ‘parallel but independent developments’²² on the other. Not infrequently, the development of commercial institutions might have followed a similar pattern neither because of sheer coincidence nor by imitation from one place to another, but because of similar economic and operational necessities.

9In this regard, a particularly relevant case is that of general average.²³ General average is the damage deliberately incurred to a ship or her cargo in order to save the rest of the cargo and the ship. The typical example is that of jettison: during a tempest, the best chance a ship often had not to sink was to throw overboard part of her cargo and/or apparel in order to lighten the hull and hopefully brave the storm. The damage was not accidental, nor did it strike masts and cargo randomly. Rather, it was deliberately inflicted on specific elements for the safety of the rest. To make the sacrifice acceptable to the owners of what was thrown overboard or cut off, some form of compensation was necessary. Hence the basic rule informing general average: any damage or loss of a part undertaken with the direct purpose of saving the rest had to be shared by all the parties interested in the voyage.

10General average is by no means the only case that could be used to illustrate the problem of ancient maritime institutions attested with seeming continuity across time and space. Shipwreck and salvage, or bottomry and loans, could have been used instead. General average was chosen because of the advantage of a more immediate link between rules and their economic implications. Also, and moreover, in general average even minor, apparently trivial variations in the application of a rule could lead to very significant divergences as to the outcome.

11The basic rule of general average is famously attested as the sea-law of Rhodes, better known as the lex Rhodia de iactu. What this lex Rhodia actually said, we do not know: nothing about it is known prior to the Romans.²⁴ Even the inscription with the general rule of the lex Rhodia found in a marble column in Rhodes’ harbour comes in fact from a Roman jurist.²⁵ Modern scholars²⁶ have little to add to what St Isidore of Seville wrote in his Etymologiae (V.17) in the VI century AD – the (original) lex Rhodia was perhaps some ancient mercantile custom of the Rhodians, whose content is wholly unknown. The words of the Roman jurist that found a place in the Rhodian harbour are more famously reported also in the Digest of Justinian: the jurist is Paul, and his words open the title of the Digest devoted to jettison, under the title lex Rhodia de iactu (D.14.2.1).²⁷ Paul’s words read as follows:²⁸

12It is provided by the Rhodian Law that where merchandise is thrown overboard for the purpose of lightening a ship, what has been lost for the benefit of all must be made up by the contribution of all.

13The rationale for the rule is then explained at the end of the next fragment in the same title of the Digest, where the same Paul explains that ‘it is most equitable that the harm be shared among those who have secured the safety of their own merchandise with the destruction of the property of others’.²⁹ The sentence is intentionally rendered in a rather pedestrian English translation, which is however closer to the letter of the Latin text. It hinges on the superlative ‘most equitable’ (aequissimum): it is not because of some – specifically Roman – rules that the loss is to be shared among all participants in the venture, but rather because it is fair to do so.³⁰ Clearly, the way the rule is then applied in Roman law depends on the specific legal framework of that legal system. But, in itself, the principle has little to do with those technicalities.

14The same rule is to be found in a variety of different contexts, some of which may not have been influenced by Roman law. So for instance this principle is found as early – and as far away – as in ancient China for damages to merchandise on boats sailing the Yangtze river, and it is not necessarily limited to sea trade, as it is also used in the Hammurabi code (ca. 1750 BC) for the damage suffered by caravans at the hands of pillagers while crossing the desert.³¹ This principle is not even limited to commerce, and some city statutes used it for fire damages to habitations. In Europe, among the places geographically (and culturally) most distant from Rome, it may be found in thirteenth-century Bergen. When a house is torn down so as to prevent a fire from spreading, the Bergen City Law (1276) provides that all the houses saved by it must contribute to its reconstruction. The rule shares all the essential elements of general average: an intentional sacrifice of a part to save the rest when faced with external circumstances (such as water or fire) that cannot be controlled. The similarity is even more striking, because the contribution is due only if the sacrifice of a house proves successful in saving the others.³²

15To explain the ubiquitous presence of general average in the Middle Ages, the German scholar Götz Landwehr wrote of ‘natural constraints’ (naturgegebene Sachzwänge) ‘leading to the development of rules that are consistent in terms of content, regardless of the respective state of legal culture and geographical location’.³³ According to Landwehr, in other words, some principles regulating commercial activities may develop in a similar way across different regions not because they spread from one market to another, but because there is little alternative to that kind of development. Operational needs, in other words, may force societies to develop similar rules for the sheer lack of viable alternatives. We do not necessarily have to accept Landwehr’s theory, but we might take it as a working hypothesis to test its limits. After all, the opposite theory – that the Roman law approach to general average spread across Europe, so that all rules on the subject derive, directly or indirectly, from Roman law – is often taken for granted yet seldom explained by scholars.³⁴

16Landwehr’s intuition does not mean that the nature of the problem dictated in itself the solution, but rather that it led to similar solutions: if the problem was the same, the solution was never identical, only (roughly) similar.³⁵ This difference might appear a mere sophism. In fact, it is significant: if it was the problem that dictated the solution, then the identity of the one would lead to the identity of the other. But it is difficult to find two customary compilations or statutes providing the same solution when faced with the same issue.³⁶ In the few cases in which this happens, the rules tend to be so remote from each other – both in space and in time – as to question the possibility that one drew from the other whilst skipping most other compilations and statutes attested in the places and the historical periods separating them. Among Landwehr’s ‘natural constraints’ we may include the risks of the sea, the legal status of merchants and mariners in overseas ports, the tasks of lading and unlading the cargo and storing it safely on board, the vagaries of the weather, the competing interests of merchants and shipmaster (and of shipmaster and crew), and so on.³⁷ Other such ‘constraints’ were less uniform, or at least more related to discrete geographical areas: for instance shipbuilding technology, navigation techniques, weaponry and ship propulsion (sail and/or oars).³⁸ With regard to them, the divide between northern and southern (that is, Mediterranean) Europe was quite significant in the medieval and, albeit to a lesser extent, the first early-modern period. It is within the constraints of those similarities and differences that we should examine the institution of general average.

3. General Average across time and space: two examples

17The basic principle behind the concept of general average is that of risk-spreading: leaving each player involved in a common enterprise to bear any loss that may result from it would be inefficient; it is more efficient to spread the risk of damage or loss to all participants, so as to lower the risk of total loss on each individual.³⁹ Spreading the risk, however, is a rather general concept which may be applied in a variety of different ways. This is indeed what happened with general average. Its concrete application did differ from place to place. Often, the reason for the different application is to be found in economic factors; other times, in cultural or social reasons. We need to be aware, however, that in most of the cases in which we reach a conclusion, we are just making an educated guess.

18Historically, the main alternative to the spreading of the risk was simply to pass it on altogether to someone else. Instead of general average, insurance. Insurance is a more refined concept, which requires a number of additional elements – some of them economic (for instance, a market with a sufficient number of players willing to underwrite a policy and sufficient liquidity to cover losses), others astride social and legal categories (such as the tolerance of usury, given the affinity between interests and insurance premium), and others properly legal (from the concept of risk transfer to the availability of legal remedies to force the insurers to pay up in case of mishap). Even in the presence of all such requirements, then, the choice of insurance over general average is not, strictly speaking, necessary. Even if it might appear more efficient from our modern vantage point⁴⁰ (so much so that in economic literature general average is usually treated as a 'precursor of insurance proper'⁴¹), this enhanced efficiency rests on a number of elements that we ought not to underestimate. To come back to Landwehr’s terminology, insurance is not an institution arising from natural constraints, but merely an alternative that looks (to us) more advanced. Even then, however, this is not entirely true, as general average is still in use today⁴² – it suffices to think of the Ever Given case (the cargo ship stranded in the Suez Canal in 2021) and of the Dali case (the cargo ship that destroyed the Francis Scott Key Bridge in Baltimore in 2024), to cite only two of the best-known incidents reported in the media across the globe in recent years. In both cases a general average was declared, giving rise to claims for several hundred million dollars.

19As said, general average is attested among nearly all societies engaged in maritime commerce: when this is not the case, the exception is more often due to insufficient evidence of its use than to clear proof of its absence. As this article uses general average as a means to a different end, no exhaustive list will be given. Rather, an attempt will be made to see how the same principle could differ in practice. To do so, we will focus on two aspects of the application of the general average principle: the contribution of the shipmaster and the valuation of the cargo. Both issues could lead to very divergent applications of the same principle informing general average – namely, that the loss or damage suffered for the common safety be shared among all interested parties. Is the shipmaster an interested party too? And, if so, should he contribute in respect of his reward for carrying the merchandise (i.e., the freight), the value of his ship, or both? Likewise the valuation of the cargo could lead to substantial differences in the outcome, and therefore in the actual contribution by the merchants. Should the merchandise be valued according to the value it had when initially purchased and stored on board, or at the value it would fetch at destination? Also, should the same criterion apply throughout the whole voyage?

4. Medieval commercial rules and their interpreters: the consent to jettison

5. The problem of Roman law in medieval and early-modern general average

58Roman law has the surprising tendency to resurface at different times over the legal horizon, especially (but not exclusively) in Europe. Today, one could easily dispense with the whole medieval and early-modern developments of general average and simply compare the Digest text that we have seen at the beginning of this article with the latest version of the York-Antwerp rules (2016),¹³⁷ to conclude that some modern commercial rules are just a re-elaboration of Roman principles. The fact that the legal terminology in use today is often deriving from Roman law lends some veneer of truth to a continuity that is more formal than substantial.¹³⁸ If we want to know what the rule was, and how was it applied, during the long centuries separating us from the Romans,¹³⁹ we need to scratch that surface, thereby removing that veneer of continuity. And here the story gets more complicated.

59One might well wonder why we need to know all that in the first place – a question that time and again has doubtlessly crossed the mind of more than a few first-year law students. The immediate answer of course lies in the enduring importance of Roman law. Behind this reason, however, there is another and somewhat counter-intuitive one: some knowledge of Roman law is necessary not to fall into the assumption that Roman law was really used so much. Presence and actual use of Roman law are two very different things.

60 Roman law is often viewed as a bridge between antiquity and modernity: this is not incorrect, so long as we study it from the viewpoint of neither of the two ages, but of the bridge itself. Looking at the great medieval civil lawyers who taught for centuries at the most prestigious universities in Europe, a first point to make is a very banal one, but no less necessary. Medieval jurists considered each fragment of the Digest as law, not as a commentary upon it. The change started with Justinian, whose legal compilation extrapolated texts – sometimes short sentences, other times lengthier passages – from jurists’ commentaries on legal issues and transformed them into legal rules. Justinian’s monumental Digest however was soon put aside even in Constantinople, let alone in the West, as far too complex.¹⁴⁰ Several centuries later, when Western jurists began to actually use it, first in northern Italy and soon then also in other European regions, they did not consider those texts as comments but as normative prescriptions – that is, texts which had themselves the force of law. The point is often taken for granted by legal historians, yet it had enormous repercussions for both the use and the understanding of Roman legal sources, signalling a profound and irretrievable rupture with the past. Even if the legal sources remained the same, the way they were used changed dramatically, to the point that one might consider the medieval approach to Roman law as parallel but thoroughly separated from that of antiquity. It is from that moment, for instance, that Paul’s opinions on jettison acquired a normative character. The consequence of this shift is that understanding what exactly the position of Roman law on the subject was matters more to a modern Roman lawyer than to a scholar interested in medieval law. For a medieval jurist, Paul’s words were the law on jettison.

61Because what Paul stated was part of the law, it was necessary to dwell on it, whether or not there was any real interest in the subject of general average. Few law teachers in a medieval university expected their students ever to encounter a general average in their professional life. But that did not make it any less necessary to study the subject: both because it was part of the law (so that it could not be skipped) and, especially, because it was possible to use those texts for altogether different purposes. The more the use of analogy in legal argumentation became widespread, the more any text could be put to good use – even for things that the jurists whose opinions had been collated in Justinian’s compilation would never have dreamt of. So for instance medieval jurists used the text of the lex Rhodia to affirm the emperor’s universal sovereignty,¹⁴¹ and even to justify taxation – levying taxes is, after all, a way of sharing the expenses incurred for the safety of all,¹⁴² and the ship had been the best image to depict the state since the times of Plato.¹⁴³

62 If we take the Great Gloss of Accursius (d. 1263), for instance, we may see that it provides a careful and meticulous discussion of the most important elements of title 2, book 14 of the Digest: who should contribute¹⁴⁴ and how,¹⁴⁵ what remedies are available to ensure this contribution,¹⁴⁶ what happens if the cargo is unloaded on lighters and some then sink,¹⁴⁷ and so on. Accursius understood the spirit of Paul’s excerpt perfectly – the question, he wrote, was not to pay for some damage to private property, but rather to compensate for the loss on equitable grounds.¹⁴⁸ The different valuation of jettisoned and non-jettisoned goods also responded to the same logic. This led Accursius to suggest (‘perhaps’) that if some merchandise arrived at destination but their market value, instead of increasing, had dropped even below the initial cost price, they ought to be reckoned after their current value.¹⁴⁹ The Roman jurists, Paul first of them, would have surely agreed. Accursius’ commentary was thorough, just as it sought to be with every part of Justinian’s legislative corpus. Thorough, but not particularly insightful, briefly summing up the most intricate points discussed in the Digest.¹⁵⁰ The issues that interested most medieval learned jurists were based on land and obligations, not on commerce – let alone maritime trade.

63If we look at other pre-eminent jurists who, unlike Accursius, did not seek to provide a commentary on each and every excerpt of the Digest but only to discuss what they thought was important in it, we will find precious little on D.14.2. A contemporary of Accursius and his colleague in the Bolognese academy, Odofredus de Denariis (d. 1265) devoted a single page on D.14.2, saying only what he thought necessary to make sure his students would understand its main points.¹⁵¹ The two most celebrated civil lawyers of the next century, Bartolus de Saxoferrato (1314-1357) and Baldus de Ubaldis (1327-1400) wrote ponderous works on the first part of the Digest (which, because of its size, was commonly divided in three volumes). But they spilled precious little ink on averages. For them, a few lines were enough. The famed fifteenth-century jurist Paulus de Castro (d. 1441) devoted a few more lines to the subject, but only because an excerpt in that title gave him a perfect occasion to discuss subtle questions of legal causation.¹⁵² The mere existence of medieval commentaries on the Roman law of general average, then, does not mean that medieval jurists were experts on the subject, and even less that Roman law was actually used to regulate the subject in practice. Medieval jurists had to comment on the views of ancient Roman jurists on jettison because, by then, those views were the law and could not be ignored altogether. But they paid perfunctory courtesy to them, acknowledging their existence and then quickly moving on.

64The interest of later jurists in the Roman law of jettison – or the lack of it – depended on the practical use of those sources. Stated otherwise, the question of whether and to what extent Roman law rules on general average were relevant in practice depended on whether and to what extent the judges were supposed to decide on their basis. The point is not as circular as it might seem. In medieval Europe, Roman law played little role in maritime commercial issues and, as recently observed by Albrecht Cordes, ‘[n]ot much more than a faint echo of the Lex Rhodia can be perceived in the [maritime] statutes of the 13^th century.’¹⁵³ The situation, however, would change during the first early-modern period, when law courts (staffed by professional, university-trained Roman lawyers) progressively imposed their jurisdiction (also) on maritime trade. The Roman law tide grew first in southern Europe, and then it progressively spread also to central and (though less uniformly) northern Europe. The growth of the jurisdiction of courts staffed by learned jurists was encouraged (or at least not opposed) by central governments, bent on imposing their power across their territories, to the detriment of older local authorities. This caused profound frictions in the rules applicable to maritime trade: law courts had to strike a balance between upholding ancient customs and applying the – Roman – law. The problem is a complex one, and it has not received sufficient attention by scholars yet. An example might help to clarify the point.

65During the first early-modern period, the scope of general average became increasingly wider.¹⁵⁴ Law courts had to take into account this without forsaking the Roman law on which they were supposed to decide. The usual way of combining these two – in many respects, opposed – needs was to rely on the interpretation of the Roman law texts by contemporary jurists. This was a standard technique in use among late-medieval and early-modern lawyers: applying Roman law texts not directly but as interpreted by some jurists of great renown. This approach became so widespread that, by the sixteenth century, the jurists’ opinion on the law often carried more weight than the law itself, especially when that opinion was shared by many jurists (the so-called communis opinio, ‘common opinion’). Some among the most important early-modern jurists on commercial law acknowledged as much expressly, among them the famed Giuseppe Lorenzo Maria Casaregis (1670-1737):¹⁵⁵

66Through the interpretation of our law professors, those texts of the Roman and Rhodian law … have been extended to any other damage whatsoever, which in similar cases was incurred voluntarily, whether on land or at sea, in order to save the goods of others

67In corroborating this statement, Casaregis cited several jurists (Peckius, Weitsen, S. van Leeuwen, Vinnius, Kuricke, and Marquart),¹⁵⁶ none of whom was Italian, Spanish or Portuguese. The selection was not casual: while Mediterranean jurists tended to be more conservative in their approach to Roman law, their northern colleagues proved more open to accepting solutions substantially different from Roman law while at the same time formally acknowledging its application.¹⁵⁷ In carefully selecting his authorities, therefore, Casaregis could prove that Roman law – as commonly interpreted among modern Roman lawyers – was not opposed to contemporary practice. Showing the consistency between Roman law and market practice was of vital importance for those jurists who had to apply the former to regulate the latter. But we need to be aware that such a consistency was often more formal than real.

68When no amount of flexibility could reconcile them, Roman law would give way to modern practice. Yet few jurists would openly acknowledge as much. A rare case comes again from Casaregis. Tellingly, however, he managed to justify as much in accordance with the canons of that same Roman law he had to push aside: the matter, he said, is only a banal question of taxonomy, where a later law takes precedence over an earlier one. So, when the Consulate of the Sea is not in accordance with Roman law, he concluded, the Consulate is to be preferred not because Roman law may be disregarded, but because the law itself provides that a newer statute is to be applied in the place of an older one (lex posterior priori derogat).¹⁵⁸ This principle was important enough to open the Theodosian Code (C.Th.1.1.1) and, especially in the works of early canon lawyers, became well established among medieval jurists, so that it appeared in the Gloss on the Digest even though it was not explicitly stated in the text of the Digest itself.¹⁵⁹ The point, however, is that the Consulate of the Sea, which Casaregis affirmed to have been included in the Genoese statutes as early as in the twelfth century,¹⁶⁰ was in fact never mentioned in the legislation of Genoa; it simply remained a source customarily applied also in that city. Its use was a fact, and it could not be avoided, but this fact had to be reconciled with a system nominally based on Roman law.

69While maritime commercial customs developed, Roman law in the background did not. The process of nominal reconciliation of modern, non-Roman practices with Roman law sources may be also glimpsed from the decisions of high courts. To spot it, though, we need to be aware of the subtleties employed by jurists. In the same city of Genoa, a famous collection of commercial decisions of its important high court (its civil Rota) was published in 1582¹⁶¹ and had a great influence in Italy and beyond. In it, an (undated) decision dealt with the liability of the insurers for the freight of the jettisoned cargo. A shipmaster had taken up an insurance policy, which covered any risk with the exception of jettison and other cases of general average. In customary Genoese practice, influenced by the Consulate of the Sea (ch. 96), freight was due for the jettisoned merchandise if the jettison happened during the second half of the voyage (and so, at a point of the route that was closer to destination than to the port of departure), while it was not considered as earned if the jettison took place during the first part of it.¹⁶² In this case, it is likely that the mishap took place during the first half of the voyage, as the decision reports twice how, because of the jettison, the freight was not paid.¹⁶³ To make up for the unpaid freight, the shipmaster then sought reimbursement from the insurers, who refused to pay. Their refusal was based on the simple observation that, if they were not responsible for the jettison, they could not be liable for its consequences either.¹⁶⁴ The point had some strength in law: if the ‘immediate act’ (actus immediatus) was excluded, the insurers argued, then all the more no ‘mediate act’ (actus mediatus) – that is, anything deriving from it – could be included.¹⁶⁵ The judges however sought to reach the opposite conclusion. To do that, some foothold in Roman law was needed. The court found it in two ways. First, in the fact that Roman law favoured a strict interpretation of the terms of the contract, something visible both in various excerpts from the Corpus Iuris of Justinian¹⁶⁶ and in the interpretation of some important civil lawyers.¹⁶⁷ Second, by arguing that freight and jettison were two different things, so that the one could well stand without the other. To press the point, the judges went as far as stating that the immediate cause of the loss of freight was the non-delivery of the jettisoned cargo at destination, not its jettison.¹⁶⁸

70Even a summary look at the Rota’s arguments, however, leaves some doubts as to their strength. First, the Roman law texts supporting the strict interpretation of a contract’s terms are far from watertight, and the legal principle struggles to emerge from the texts quoted.¹⁶⁹ The omission of many other more relevant texts in Justinian’s compilation does not appear fortuitous, as it might not have led where the judges wanted to go. Also, the choice of the civil lawyers quoted might not be entirely representative of the common opinion among the Roman law doctors, as it would have been possible to find many famed professors arguing for a different solution. Finally, one does not need to be a refined Roman lawyer to see through the last and main point on which the Rota reached its conclusion: the fact that jettison and freight are two different things has little to do with the causal relationship between the occurrence of the first and the non-payment of the second. No Roman law text, civil law jurist, or civil law court is known to have conditioned a causal relationship on the subordination of one element to another.¹⁷⁰ Arguing that the loss of the freight was directly imputable to its non-delivery and not to the fact that the cargo had to be thrown overboard is specious at best, even if one were to apply chronology to assess causation – an approach, by the sixteenth century, long discarded among both jurists and courts throughout Italy.¹⁷¹

71The insured, in short, was entitled to recover from the insurers the freight that the merchants had not paid because of the jettison, even if the policy had explicitly excluded jettison from the risks insured against. This, according to the Court, was the Roman law position on the matter. To make sense of this liberal use of Roman law, we need to look at contemporary commercial practice. In the early-modern period, merchants would no longer accompany their cargo on board. While legal literature still commonly referred to the representative of the merchant – the ‘sopraccarico’, who was responsible for (sopra) the cargo (carico) – as being present on board,¹⁷² in practice neither merchant nor sopraccarico routinely followed the merchandise. As a consequence, the only witness of the shipmaster’s actions was his crew. This left the shipmaster in full control over the narrative of the events. With little risk of being contradicted, the shipmaster had all the interest in portraying any accident happened to his vessel as a case of general average, so as to receive a substantial contribution from the merchants.

72Roman law restricted this contribution only to the case where the damage to the ship was voluntarily incurred also for the safety of the cargo. In any other case, the loss of riggings, masts and yards was on the ship alone (D.14.2.6). Paying lip service to Roman law was not difficult: it was sufficient to describe the damage to the ship as done for the safety of the cargo even in the most unlikely cases. So a typical scenario in which masts broke down during a tempest could be easily transformed into a heroic struggle to save the merchandise,¹⁷³ and a ship that began to sink as soon as she left the port could be said to have rescued the cargo on board by rushing back to to the pier.¹⁷⁴ Soon, the description of the mishap began to follow pre-determined patterns which, it was observed, look like a ‘formulaic standard specifically designed to trigger a general average declaration’, clearly ‘designed to counter the most predictable objections to the sacrifice.’¹⁷⁵

73The broader the general average, the more it became possible to claim from the merchants. In seventeenth-century Leghorn, for instance, a shipmaster could receive compensation for the most disparate expenses, which had patently little to do with the rescue of ship and cargo from danger: seamen’s wages, costs incurred for the arrest of the ship by public authorities, interests for money borrowed during navigation, divers’ fees, local taxes and even bribes routinely paid to officials.¹⁷⁶ The situation in Venice was not dissimilar.¹⁷⁷ Even though those expenses were described as being incurred for the safety of the cargo, few merchants would have bought it. Yet, pay they did: by and large litigation did not focus on the merchants’ refusal to pay, but only on the amount of their contribution.¹⁷⁸ Progressively, therefore, the criterion for general average shifted from common safety to common benefit, although formally it remained anchored to the Roman law criteria. This change was hardly the peculiarity of some Italian markets. It is already attested in important northern centres such as Antwerp during the sixteenth century,¹⁷⁹ and even earlier in Bruges.¹⁸⁰ Just as in Italy, also in Antwerp merchants would typically accept the ever-increasing scope of general average, bringing a lawsuit only when the requested contribution was exceptionally high.¹⁸¹ In seventeenth-century France the same trend was so advanced that it even led to a wholesale reconfiguration of the subject. Instead of listing what operational costs could be recovered by way of general average, the Ordonnance de la Marine of 1681 did the opposite, barring only one kind of expenses (guns and ammunition) and allowing any other (III.8.11).

74If, on a formal level, general average remained unchanged, substantially it became increasingly akin to hull insurance – without the restrictions provided for that instrument. Admittedly, the great increase in the tonnage of most ships during the early-modern period often made the cargo more valuable than the vessel that transported it. Still, the merchants were no longer bearing the risk of the sea only for their goods, but largely also for the ship. This of course was never stated openly, but it was clear to all. To understand the point, once again, we need to look outside the law. In early-modern Europe, the growing competition in maritime commerce between states led to an increasingly protectionist attitude in favour of national shipping industries.¹⁸² An easy way to do that was to shelter shipowners from excessive expenses that would have crippled their business, shifting part of those expenses onto the merchants. Challenging the account provided by the shipmaster, as we have seen, was remarkably difficult for the merchants. Instead, they soon realised that the only efficient response they had was to accept the situation, and make sure that someone else would pay for their ever-increasing contributions: the insurers who underwrote their cargo policies.¹⁸³ Thus, this progressive enlargement of the scope of general average ultimately transformed its very function, from a risk-spreading mechanism to a risk-shifting arrangement in which each group taking part in the venture was able to pass on a substantial part of their risk: the shipmaster onto the merchants, and the merchants onto the underwriters.¹⁸⁴

75It was only towards the end of the eighteenth century that the excessively broad scope of general average became a hotly disputed question. By then, insurance had become to a large extent institutionalised: insuring the cargo had become almost the norm rather than the exception, and the insurers were professionals specialising in the field – often, powerful syndicates such as Lloyd’s or corporations, which began to put pressure on governments to reduce their exposure to general average claims. Moreover, widespread insurance also meant that many merchants and shipmasters were themselves insured: in such a case, both the merchants whose cargo had been jettisoned and the shipmaster who lost some equipment would be overcompensated if they could recover under their policy alongside the contribution due to them under general average.

76In his ponderous treatise on maritime law, published in the 1780s, to introduce the subject of general average Ascanio Baldasseroni noted how:¹⁸⁵

77Infinite … are the abuses introduced in this subject, so that seldom the damages suffered by a ship, whether for her own vice or for a mishap during the navigation, were not ingenuously dressed up as a consequence of a voluntary act aimed at the common safety. This way, as soon as shipmasters and shipowners have a ship at sea, they deem it lawful to consider her as a perpetual asset, and whatever the use [of the ship] they make, and though the many repairs leave not a single plank from the initial frame, they always find a way to allege that the replacement of the parts has preserved the whole. Thus, what is not true becomes in their own interest an actual fact.

78A similar practice, continued sarcastically Baldasseroni, is more reminiscent of the legendary ship of Theseus, kept intact for a thousand years by surreptitiously replacing the old parts. ‘But in our days such miracles happen only to the detriment of the insurers, who are forced to keep alive vessels, either by way of general average befalling on them, or by way of particular averages [i.e., insurance proper] when they take up the risk in the hull and appurtenances.’¹⁸⁶

6. Conclusion

79Towards the beginning of this article Landwehr’s idea of ‘natural constraints’ was discussed as a working hypothesis to explain the seemingly ubiquitous presence of some legal institutions underpinning segments of pre-modern commerce, as well as their surprising resilience. This of course has nothing to do with universal concepts such as lex mercatoria and other such myths. Rather, it is an effort to explain the presence of the same legal principles in many heterogeneous societies without postulating a common normative origin.Landwehr’s intuition is fascinating, though of course not sufficient to predicate the universality of any legal institution a priori. More important is the fact that, no matter how widespread or indeed ubiquitous some legal principle might be, its practical application tends to be highly variegated, to the point of making it very difficult to reconstruct its origin and trace it back to a single legal system. The number of variables that may change is large, and the rules that were applied in practice are far more dependent on those variables than on the general principles. The only meaningful comparison that may be undertaken ought to take place at the lower level of those practical rules, not at the higher one of those principles, no matter how inspiring that approach might appear.

80To make sense of those rules it is essential to look at their context. This is trite knowledge, but it is worth repeating it because there may be aspects of that context influencing the rule in ways that would not strike as obvious. The analogy between camels and ships worked out by Iraqi scholars is one such case. Looking at the context requires awareness of changes that happen without leaving clear traces, but which are nonetheless important to understand the working of a rule. Imagining that refined written elaborations of older oral customs left no mark on their practical application would be wrong, just as it would be disingenuous to imagine that those re-elaborations found punctual application in practice for the simple reason that they were now the rule. By the same token, if we look at early-modern Europe, it would be a fallacy to take for granted that Roman law was applied only because in several jurisdictions lawyers and judges alike were supposed to rely on it. Invoking Roman law in order to reach a different solution was a refined way to square the circle, maintaining (non-Roman) practices within a legal environment imbued with Roman law. For our purposes this means that, when a legal principle is reminiscent of Roman law, it does not necessarily mean that it is derived from it. And, even when this is indeed the case, Roman law itself may not be sufficient to study it. Discarding Roman law entirely, however, may lead to even bigger problems than assuming its wholesale application. It needs to be handled with care.