Revista Ítalo-Española de Derecho Procesal
pp. 165-187
Madrid, 2022
DOI: 10.37417/rivitsproc/1805
Marcial Pons Ediciones Jurídicas y Sociales
© Jordi Nieva-Fenoll
ISSN: 2605-5244
Recibido: 23/07/2022 | Aceptado: 03/11/2022
Editado bajo licencia Creative Commons Attribution 4.0 International License.

REQUIEM FOR THE BURDEN OF PROOF

Jordi Nieva-Fenoll*

ABSTRACT: The burden of proof, a notion specific to the medieval Roman-canonical process but alien to the four Roman procedural systems, ought to have become obsolete with the introduction of the free assessment of evidence. However, doctrinal and jurisprudential inertia in the use of traditional concepts, as well as the conservation of biphasic processes in legal systems of Anglo-Saxon origin, including the Roman-canonical process, have favoured the persistence of a notion that, when observed objectively, has ceased to have any legitimate practical value in current judicial processes.

KEYWORDS: Free assessment, summary judgment, standards of proof, formulary system, Roman canonical process.

Index: 1. INTRODUCTION.— 2. THE SUBJECTIVE BURDEN: A LIVING CORPSE.— 3. THE OBJECTIVE BURDEN: A JUDICIAL ACTIVITY WITH AN IMPROPER NAME.— 4. THE BURDEN OF PRODUCTION: A USEFUL INSTITUTION IN AN ANACHRONISTIC PROCESS.—5. THE BURDEN OF PERSUASION: AN IMPOSSIBLE DESIRE.

1. Introduction

Four years ago, with overwhelming fear and respect, I first suggested 1 that the institution of the burden of proof should be abandoned in both its objective and subjective versions. This was by no means an original conclusion; more than a century ago, Kohler 2 and Bar 3, among others, had already pointed in that direction, an unavoidable consequence of introducing the free assessment of evidence. Even Rosenberg 4 came to this conclusion about subjective and—eventually—objective burden, although the latter was alien to his intentions 5. These abolitionists were not lacking arguments; as we shall see later, the burden of proof was indeed incompatible with the system of free assessment of evidence 6.

It was perhaps due to the influence of Wach’s 7 erroneous but forceful and harshly formulated opinions of the time, and most probably due to the simple linguistic and behavioural inertia of professors, lawyers and judges, that the “burden of proof” has survived, rather surprisingly, as an institution—or perhaps rather as an expression—in case law and procedural law manuals. Although the doctrine continues to refer for the most part to the burden of proof in its most primitive sense, the subjective, or, if objective, the “subjectivisation” of it through a very confusing expression—”risk sharing”– 8, what is being used in the courts is not really the burden of proof. However frequently mentioned, its genuine content is not being applied in this same judicial praxis. Instead, guidelines on the assessment of evidence are being used, covered under the solemn mantle of “the burden of proof” 9. Remarkably, this kind of ghost expression has remained with us, erroneously referring to something that no longer exists, a phantom in the attic. Whether for the sake of convenience or tradition, it serves no purpose.

In the following lines, I will try again to defend this forceful appraisal, although this has already undergone full written confirmation by Michele Taruffo 10, and on the subjective side only by Jordi Ferrer 11, as well as—years before I formulated it—by Barbosa Moreira 12, though only in a minimal way, and by Luca Passanante 13, among others, on the objective aspect. But the publications that have emerged so far on this abolitionist issue, writings that still defend the validity of the concept, have been extraordinarily cautious 14.

Although it may be the simplest and most commonplace thing to do, it is impossible to support the existence of an institution simply by resorting to arguments ad antiquitatem, ad populum and ad verecundiam, three fallacies that claim to confirm that a conclusion is true simply because “since ancient times” it has been defended by “everyone”, including “authors of recognised prestige”. Things are not that easy. A conclusion is scientifically sound when its correctness can be epistemically defended, and when empirical data corroborating its presence appears, making that first step obvious. For example, res judicata exists because legal systems prohibit the repetition of judgments, which is perfectly tangible in practice, and not because doctrine or case law speak of it. On the contrary, many legal systems speak of the burden of proof, but the empirical significance of their words is null, beyond a simple indicative orientation, as we will see later. In any case, a legal institution cannot be sustained on faith. It must be apprehensible. Otherwise, it becomes a Russell’s teapot 15, a surprising analogy that can even be formulated through the traditional logic of the burden of proof, although it has nothing to do with it: the existence of a concept must be proven by the one who affirms its existence, not by the one who denies it. Thus, when one tries to prove that the burden of proof exists, it is shown to be an unnecessary concept. However, it is also possible, at least in this case, to prove the non-existence of the notion in a system of free assessment of evidence, suggesting that the logic of the burden of proof itself is merely apparent.

2. The subjective burden: a living corpse

The distinction between the objective and the subjective burden of proof is alien to Roman-canonical process 16. However, the subjective burden of proof is the only aspect of this institution which, despite having been scientifically isolated in the 19th century in a somewhat artificial way 17, coincides with its original meaning. It starts from a very rudimentary basis: every litigant in the process must provide evidence of what they claim, otherwise they will lose. So rudimentary was the idea—not the institution—that it can even be found before the history of Rome in the Code of Hammurabi 18, which served nothing less than the death penalty for those who did not bring their evidence to the trial 19. It is curious that even with such a radical provision, no one seems to have claimed that the burden of proof was born in Hammurabi’s time. On the same basis, albeit heavily influenced by medieval doctrine, the classical authors 20 closest to our times affirmed that the plaintiff had to prove the facts constituting the claim, and the defendant the impeditive, extinguishing and excluding facts 21.

Contrary to what has been said, the notion of the burden of proof does not come from the Roman process in any of its four historical phases. Kaser 22, though he cites the institution, tiptoes over it. He practically excludes it in the process of the legis actiones 23, and mentions it for the first time in the period of the formal process, though not as an institution applied by any iudex in the apud iudicem phase, nor even by any praetor in the in iure phase, but simply as an expression of a basic idea: the logical thing is that each party proves what it asserts. This does not exclude—as Kaser explicitly states—that the iudex, or the iudices, could take advantage of the evidence of the opposing party to support the opposing party’s position 24, as was logical in this process, something that was usually decided by juries 25 without any control over a supposed application of the burden of proof, since such control simply did not exist. This is a curious first reference to the so-called principle of acquisition 26.

One only begins to detect something similar to the institution—created only during medieval times in my opinion—in the Roman classical process of cognition, or cognitio extra ordinem. Kaser 27 refers to the interlocutio which would have dealt—again, according to the author—in that process with the question of the burden of proof. However, it is difficult to say whether the fleeting interlocutio to which Kaser 28 refers was concerned with the admission of evidence—as it seems to be 29—or with a supposed application of the burden of proof, which is perhaps less likely, as I will explain below. Finally, in the post-classical process, which is already directly inspired by the Roman-canonical process, Kaser 30 recognises that the institution becomes blurred.

Kaser’s account is surprising: the institution is barely glanced over, suddenly appearing and then almost vanishing. However, there may be a good explanation for this kind of “burden of proof” escapism. In my opinion, it is not that a so-called institution called “the burden of proof” ceased to be used. If one reads the whole of Title III of Book XXII of the Digest, they will see that these passages speak of its title, probationibus et praesumtionibus, that is, of proof and presumptions, and not the burden of proof. To put it more simply, it speaks of the free assessment of evidence, which was the system of assessment of evidence in force at that time in Rome. The text merely establishes a series of guiding rules for free assessment in specific cases, citing the most typical clues to be found in some of the most frequent processes of the time, as it does with regard to documents in Title IV of Book XXII, and even with witnesses in Title V. Some of these indicative statements in fact served as a basis—or pretext—in the late Middle Ages to create the legal rules of evidence 31, but originally they were not at all legal rules of evidence; in Rome, at least in classical times, the system of free assessment always prevailed.

Although it may seem otherwise in a more hurried or excessively literal reading, the Roman jurists did not actually state in these paragraphs who has to provide proof. This was a conclusion that was drawn in medieval times with what in my view was an ultra-literal interpretation, something very typical of that period of scholasticism. In reality, what those jurists were doing was merely setting out the most characteristic indications of the most difficult processes, applying the essential basic logic already mentioned to them, and thus concluding that whoever affirms something must prove it. Consequently, Title III together with IV and V of Book XXII of the Digest is not a small monograph on the burden of proof. It is a simple treatise providing guidelines on evidence, like many other more extensive ones that were written later 32. This tradition has survived to the present day 33.

What Kaser said had become blurred was not the burden of proof. Instead, with the occurrence of many different cases over several centuries, the indicative rules guiding the judge contained in Book XXII of the Digest, which simply guided the judge on what evidence to use in each process as well as who was most likely to have that evidence, became a constant question of nuance.

These are precisely the rules—always admonitory in Roman times as free assessment was in force—that in the late Middle Ages were turned into rules of legal assessment by employing the scholastic method. They were often based, to be clear, on the literality of authoritative texts 34. In fact, when Kaser 35 claims that only Justinian tried to return to the tradition of classical period in this matter by supposedly recovering the burden of proof, what the emperor actually did was recall the basic mnemonic rule that went back, as we have seen, to at least the time of Hammurabi: he who asserts something must prove it 36: quia semper necessitas probandi incumbit illi, qui agit, or, as Celsus wrote at the beginning of the second century, quod qui excipit, probare debeat 37. Notably, the first passage reproduced above does not speak of onus, but of something quite different, and certainly not an obligation: necessitas.

Therefore, proof of the existence of the institution we know today as the burden of proof is missing in the Roman sources. In fact, the notion was doctrinally created in the medieval process of the solemnis ordo iudiciarius 38. In that process, it was the result of the gloss and commentary of the Corpus Iuris Civilis, a procedure that was conceived after a very cumbersome preliminary phase (praeparatoria iudicii) 39 that attempted to verify the subsistence of the dispute and the inexistence of procedural defects, exceptiones dilatoriae. A second phase began with the litis contestatio 40, i.e., the beginning of the cum testes process, or in other words, the proposition and practice of evidence.

The fact is that this second phase began with a brief claim and an even briefer reply in which the plaintiff’s request was simply denied 41. After both parties had taken the oath of slander—something that could end the proceedings unfavourably for the party who did not take the oath 42—the most important phase of the proceedings began. The plaintiff had to formulate his positiones, that is to say, their assertions of fact, to which the defendant replied with their responsiones, which were also assertions of fact formulated in their defence 43.

The list of positiones and responsiones formed the roadmap for the subsequent course of the proceedings 44, namely, the taking of evidence from those positiones and responsiones. Once they had both been formulated without any solution of continuity or express judicial decision—interlocutio probationis only occurred at a very early stage of the Roman-canonical process before disappearing 45—each litigant had to offer proof of each of the facts on their list 46 and the burden of proof was thus shared 47. If the plaintiff did not do so from the outset, the process had to be concluded 48 unless, to cover their evidentiary deficiency, they requested the oath of the opposing party, who usually had to take it in order not to lose the process. The oath is a nonsensical remnant of old ordeals 49 that is surprisingly still present in several countries 50. If, on the other hand, the plaintiff offered proof but the defendant did not offer their own, they were considered to have confessed and the proceedings were also concluded. Thus, the burden of proof, the onus probandi, was born, and it was taken into consideration, as can be seen, before the taking of evidence. Therefore, if the question is whether at this stage a litigant could lose the case because they did not meet the burden of proof of the facts they allege, the answer is undoubtedly yes.

At that point, the role of the burden of proof—in a clearly subjective version—had almost disappeared, but it still had a mission at the end of the process, something that has probably misled the doctrine by mixing up the burden of proof with the assessment of evidence in that same medieval process. At that time, let us remember, lawyers defended themselves in proceedings in the aforementioned preliminary phase (praeparatoria iudicii), formulating, above all, those dilatory exceptions with the aim of finding some procedural defect that would paralyse the plaintiff’s claim. They also analysed whether both parties agreed, which could favour not only acquiescence or waivers, but also settlements. Once this phase had been overcome and the litis contestatio had been reached, the lawyers’ task consisted of either complaining that the positiones had been formulated in an unclear manner 51—the antecedent of the exception of defect in the way of proposing the claim 52—or that the responsiones were in fact confessiones 53, which made the proof unnecessary. But once both lists of facts had been established, the litigants, as already mentioned, offered their proof, thus complying with the onus probandi. At the same time, in that precise moment, efforts were also focused on objecting to witnesses and documents.

The reason for these objections was very clear. At that time, judges did not listen to witnesses 54 and sometimes did not even read documents as they were often relatively illiterate 55. Instead, perhaps because of the influence of Germanic law and its very frequent use of witnesses to confirm—or attest to—legal acts 56, their credibility was not assessed by listening to them, but simply by ascertaining that they were willing to confirm under oath what the plaintiff or defendant said, which was entirely logical from a purely religious point of view. The same thing happened with public documents: a notary was willing to confirm their veracity with his faith, another way of taking an oath.

However, as I have said, the effort in this phase also lay in objecting to witnesses and documents, given that neither one nor the other could be freely assessed. There was no way of doing so since the oath was imposed; the proof that had remained for each party was simply added up, and the litigant who obtained the greatest amount won the process 57. In other words, the winner was they who had better fulfilled their burden of proof by placing a greater weight of evidence on their respective plate in an imaginary procedural scale of justice 58. The distinction between plena probatio and semiplena probatio (also called probatio summaria) was of great importance in this situation since the former had greater legal value than the latter 59. Plena probatio was usually fulfilled with a pair of witnesses, a public document or an iuramentum veritatis 60–much more infrequent than slander oath, which was systematic and did not have full probative value–so that someone with 10 witnesses and one public document possessed six plenae probationes. A single witness or a non-public document was reduced to the value of semi-full proof, though this was variable 61, about half. Thus, a litigant with 10 witnesses, one public document and one private document had six full items of proof and one 0.5 half-full proof, thus totalling 6.5. If their opponent had managed to add 12 witnesses and two private documents, he had six full items of proof and two (0.5+0.5) half-full proofs, i.e., a result of 7. In the production of evidence, the judge naturally did not have any role whatsoever, except in the iuramentum veritatis, which could be ordered by the judge in spiritual and matrimonial proceedings 62. They always judged secundum allegata et probata (partium) 63 in a full adversarial manner. Although judges could assess the evidence in a reasonably free manner 64, this was not the norm, and instead the mathematical system already described was applied. This absurdity at the time was not only the result of the aforementioned influence of Germanic law but also most likely because of mistrust in the role of judges: they were not independent, being appointed by nobility or royalty 65, and were often poorly educated in law, as already mentioned.

All of the above ceased with the reintroduction of the free assessment of evidence in the 19th century, a true procedural enlightenment and a return to the Roman past in which the burden of proof no longer existed as a mere mnemonic description of something that was simply logical—whoever asserts something must prove it 66—for the simple reason that it is not possible to initiate, maintain or win proceedings in a vacuum.

Before continuing, reference must be made again to an important statement in the Digest: ei incumbit probatio qui dicit, non qui negat 67. The phrase in question must be put into context in order to be properly interpreted. First, note that the Digest in that passage does not speak of onus probandi, as has sometimes been said in quoting the same sentence, but simply of probatio. This is already a remarkable fact. It is not until a few numbers further on, in the title itself, that Paulus, who was also the author of the first sentence, pronounces the words onus probationis 68, though it does not seem that he is referring to any institution, or even creating one. The word onus is very frequently used in the Digest, referring to many different obligations, responsibilities and even missions. But if one examines the whole of Title III of Book XXII, one sees that what these passages do is simply establish, as already mentioned, a list of indications for different processes, stating who is most likely to have proof, all with the intention of helping the judge by providing them with a guide, albeit with admonition. It does not give birth to an entire institution that would supposedly have been important for a Roman jurist. For example, when the Digest constantly refers to res iudicata 69, actio, exceptio 70, interdictum 71 or appellatio 72, it devotes whole passages to them that are absent in the case of the burden of proof.

All this contrasts extraordinarily with works of medieval law in which onus probandi is not only very present 73, but also not infrequently dealt with at the beginning of the explanations on proof 74, coinciding with the temporal place reserved for it during the process at the beginning of the litis contestatio. It seems that medieval jurists, relying on a scholastic reading that was always exaggeratedly literal of the Corpus Iuris Civilis, created an institution that made sense in their time, but which had not really existed in Rome. This influenced more modern doctrine. People often try to identify fact or logic in the present by looking for them in the past, ignoring the fact that the past must be interpreted in its own context, without help from the future, which is completely alien to it. It is possible to make the history of law travel forward in time to the present day, but trying to do so in the other direction, though highly evocative, is like trying to identify European legal institutions in traditional Amazonian law; there will always be elements that resemble each other, but they will not usually be common.

We arrive in the 18th century with onus probandi, an institution that existed in the evidentiary system of the time: the legal system. Suddenly, precisely to break with the legal system, someone suggested copying the English jury, giving rise to the struggle for the reconquest of the free assessment of evidence, which had been cornered in the late Middle Ages by the scholars of Bologna. That someone, as is well known, was Jeremy Bentham 75.

Bentham, in a similar vein to Beccaria 76, had suggested that judges should see and hear the witnesses 77 and judge according to their intime conviction 78, as that was the way juries did in his country. Bentham was first heeded in France 79. Consequently, it was no longer necessary to add up witnesses, but rather to listen to them. It was no longer necessary to add up documents, but rather to read them. Thus, what two witnesses said could no longer be worth more than what one witness said because it was now dependent on the credibility that the judge attributed to them. As the evidence was freely assessed, it was no longer possible to decide the case by looking at which party had met its burden of proof; the evidence provided by both litigants was no longer simply added up separately, taking for granted that what was provided by each party undoubtedly benefited them. It was now possible that evidence provided by one of the parties could benefit the other, which not only inaugurated the presence of Chiovenda’s 80 so-called principle of acquisition, but also rendered the application of the logic of the burden of proof useless, at least in its subjective aspect. Evidence was to be freely assessed, irrespective of who had provided it. The absurd assumptions as to the fulfilment of the burden at the beginning of the trial were no longer established. Evidence was to be assessed as a whole (§286 ZPO) 81 and facts that a judge was able to establish as true in the light of evidence were considered proven. It was the same thing the English juries did, but it motivated the judgment, and did not leave everything to the best of their knowledge 82, or intime conviction, as the French translated it.

All this rendered the “inversions” of the burden of proof, the “lightening” of that same burden 83 and the “evidential facility” 84 absurd. This, by the way, was the initial basis for the classic distribution of the burden of proof as well as, of course, that nostalgic modern formulation, the so-called “dynamic” burden of proof 85, all of which are reactions of doctrine and case law that tried in vain to preserve the validity of an institution that had ceased to make sense, completely altering its original approach. Such notions—inversions, lightening, ease, dynamic burden—are nothing more than situations identified by the majority of case law in which it is foreseen that one of the parties, usually—although not always—a vulnerable party who finds it difficult to defend themselves in the proceedings because of a lack of easy access to the evidence. Thus, their version is given initial credibility, even if it is supported by little to no evidence, and the opposing party is warned that they must make a greater effort than usual to prove what they say or disclose evidence available to them. If they do not do so, even if it is not said in such an icy or unpleasant way, the process will be decided against them as it will be understood that they are hiding evidence because it would favour their opponent. In other words, the evidentiary deficiency of a litigant who could easily obtain such evidence is interpreted as manipulation of reality to procedural advantage, or hiding evidence. Consequently, they are condemned on the basis of what is a simple indication of concealment. All this is not about burden of proof, however, but rather the assessment of a prima facie case. It is clearly about the free assessment of evidence.

In conclusion, the free assessment of evidence, with its inherent principle of acquisition, renders the study of the subjective burden of proof obsolete. It must therefore be considered doctrinally superseded. Let us now explore whether the other meaning of the burden of proof, the objective, can retain some validity.

3. The objective burden: a judicial activity with an improper name

In the judicial process of the solemnis ordo iudiciarius, as we have seen in the above, considerations of assessment and distribution of the burden of proof were mixed in at least two moments. At the beginning of the process, when those who did not offer proof, or all the proof they offered had been objected to, they were made to lose the process. It occurred at the end of the process, when the contribution of each party was “weighed” using the oldest known evidential standards: probatio plena and semiplena probatio. This activity of weighing the evidence, which was typical of the legal evidence regime, replaced the free assessment of evidence. In fact, it could be said that this “weighing of evidence”, or “legal assessment of evidence”, was not really an assessment at all, but rather a simple application of the burden of proof. Each party had met or failed to meet its burden to a greater or lesser extent. Depending on the arithmetical result, a decision was made. While quite absurd, as we know, this was very simple to apply in practice, and that is why it was successful. In the end, the courts, as a pure and logical defence against their workload, look for the simplest solutions that involve the least amount of effort, as long as they are, at the same time, apparent.

With this explanation, it is understandable for confusion to arise between concepts and the emergence of the objective burden of proof. It appeared in the doctrine in criminal proceedings 86 with the obvious realisation that the subjective burden of proof cannot be applied in such proceedings since the defendant is not obligated to prove anything and in fact the public prosecutor does not even have to prove the accusation. Contrary to popular belief, public prosecutors simply contribute to the clarification of the facts by ascertaining the prosecution’s, and also the defence’s, evidence related to the crime that is the object of the proceedings. In this way, the work of the public prosecutor is alien to the institution of the subjective burden of proof. Their mission is simply to collaborate in establishing the truth. As has long been peacefully asserted in German doctrine, public prosecutors are not a party to the proceedings 87.

Glaser 88 considered that another variety of the burden of proof persisted, however: consideration of the old standards for ascertaining which facts were proven and which were not. Reaching those evidential levels in order to establish guilt in criminal proceedings achieved a new standard, the probatio plenissima 89, superior to probatio plena, which, in part, tried to reflect in England the standard of beyond any reasonable doubt 90. It is very complex to put into practice, however, and in fact was probably the first sign of judicial liberation from the ties of the legal system of proof in favour of the free assessment of evidence, at least in part. Glaser 91 called this consideration of the determination of evidentiary levels or standards the objective burden of proof. It was, after all, the final stage of consideration of the medieval onus probandi in the legal system of proof. When the litigant provided evidence, thus meeting the subjective burden, it determined whether they had done so sufficiently.

In the legal system, this question was answered by adding up and weighing, so it would have been easy to confuse it with the subjective burden of proof, given that both were oriented towards the same objective: to decide the winner of the trial automatically. But when we move from the legal system to the system of free assessment, everything changes. It is no longer weighed, but balanced. The judge must evaluate the clues offered by the evidence without being bound by comfortable automatisms or prejudices in order to search for material that will allow them to justify why they believe an evidential result to be credible. However, this is no longer burden of proof, but the assessment of evidence itself. It is simply the assessment of that evidence that a judge will carry out simultaneously while perceiving the results of the evidence. It is not possible to perceive without assessing, and what is not assessed is overlooked, as cognitive psychology confirms 92.

Calling this process the “burden of proof”, although it has an obvious historical explanation, is currently misleading and, above all, highly disorienting. For those unfamiliar with the functioning of the old medieval process, it is very easy to become confused about the content of the notion.

A different question is whether assessment should be guided by the guard rails of evidential standards–semiplena probatio, plena probatio, probatio plenissima–as the system of legal evidence did, or whether it should be left completely to the discretion of the judge, as the system of free assessment dictates. In this sense, the attempts to redirect this matter to a logic of standards 93 have either been very polemical 94 or simply refer to phrases that attempt to guide mostly judicial bodies that do not motivate: the jury 95. The guiding effectiveness of these fine phrases in conveying their content–probable cause, preponderance of evidence, clear and convincing evidence, beyond any reasonable doubt—is more than debatable and, in the case of the jury, impossible to analyse a posteriori in the absence of motivation in the jurors 96.

However, it may be useful to apply the method on which the configuration of these standards is based, which is fundamentally inductive probability 97. It is an excellent way of assessing evidence by not focusing all activity on a single hypothesis, but also on the other hypotheses that have emerged in order to rule them out. All of this entails an inevitable subjectivity in the formulation of the hypotheses and even in the evaluation of the result. Having this subjectivity channelled through a method is not something to be rejected, but rather desirable. A second question would be how to reach evidential thresholds based on this method. Given its inherent subjectivity, the calculation of these thresholds is either radically objectified, as in the regime of legal proof, or it becomes nothing more than an impossible desire. It is not really feasible to soundly defend that one event of the past being investigated in the process, a fact subject to proof, is with total exactitude more or less probable than another. Since the judge has not witnessed those events or knows all the circumstances, the conclusions are no more than hypotheses. It is not like proving the existence of a black hole or the efficacy of a drug. In the process, we can only reconstruct traces of reality but never confirm them beyond a reasonable doubt, no matter how often the phrase is repeated. What we can do is try to minimise these doubts, and this is what this standard asks of us. Reconstructing reality to this very fine point is not really feasible, however. In the process, we will always have a version that has been proven as if we were walking on hot coals. We will be able to explain how we have cornered the inevitable doubts about a version of events, but like a historian 98, we will never be able to claim that this, and only this, is what happened. While disappointing, as humans we have limitations, and this is one of them. Perhaps in the future we will find a way to confirm legal facts with the same efficiency as corroborating the existence of the Higgs boson. For the moment, it is science fiction.

4. The burden of production: a useful institution in an anachronistic process

It comes as a great surprise that when we look at the Anglo-Saxon systems, which are different from each other but with much in common, we again come across the burden of proof. There are two varieties: the burden of production and the burden of persuasion, in addition to the burden of allegation, which evokes the old system of positiones and responsiones, something we will leave aside for the moment. It is also surprising to see how the burden of production coincides with the subjective burden of proof 99 in the version influenced by the principle of evidential facility/vicinanza probatoria 100, while the burden of persuasion coincides with the objective burden 101. What happened here?

It is a simple but little-acknowledged fact that not only did English jurists study Roman law in depth as a priority, at least until the end of the 18th century 102, but English law also considered the Roman-canonical process of the solemnis ordo iudiciarius 103. Although it followed partly different paths, above all as a consequence of the trial phase before juries, in reality the procedural structure of the Anglo-Saxon systems divided into two phases, pretrial and trial, are strongly reminiscent not only of the praeparatoria iudicii and the litis contestatio of the medieval process shaped by glossators and commentators of Bologna, where Englishmen also studied in the Middle Ages 104, but also of the phases in iure and apud iudicem of the old Roman formulary system. In fact, the in iure phase was held before a praetor, and the apud iudicem was usually held before juries 105, not unlike the pretrial (before a judge) and the trial (before a jury) in the US. The difference between barrister and solicitor is also much more faithful to the procurator/advocatus model in the English system, though this is a different issue.

The current Anglo-Saxon-inspired process has an extraordinarily old structure that its users seem to find useful, if not highly debatable. It all begins with a very long 106 pre-trial phase—just as praeparatoria iudicii were also very long—in which the aim is above all to negotiate. In the solemnis ordo iudiciarius the aim was likewise to check whether the dispute still existed. But in this phase, there is also something of the old preliminary phase of the litis contestatio: the positiones and the responsiones, that is, the claim and the defence, that the judge tries to understand requires the parties to comply, naturally, with the burden of production, or in other words, with the subjective burden of proof. And indeed they do, otherwise it is possible for the judge to say that the case has no prospects of success for the party who has not fulfilled their burden, according at all times to the judge 107. They can issue a summary judgement against the non-compliant party to avoid the trial 108. Thus, parties go on and on, producing more and more documents, summoning witnesses and making them sign affidavits in the purest medieval style of ratifying witnesses without seeing or hearing them. In this system, therefore, onus probandi, the subjective burden of proof, or the burden of production, makes perfect sense. Or does it?

We find ourselves with a system that theoretically uses the free assessment of evidence, but has retained absolute inquisitorial judicial powers in the pre-trial in a supposedly adversarial process. What is the explanation for all this? It would take a long time to specify, and merits further research, but of course fear of the unpredictability of the jury 109, and of course of summary judgment 110, have played an important role. However, there is a much more contemporary logic to the way the process has been configured. Judicial process is a public service that does not fit well with neo-liberal logic, which aims to privatise everything, including litigation, so that the most powerful do not get placed on an equal footing with others in the eyes of a judge. The least powerful would end up suffering a terrifying inequality in an area where no one is looking out for them: alternative means of conflict resolution. Thus, ineffective mediations arise, their sole purpose being to wear down the weak with time and money, with arbitration in the hands of arbitration institutions influenced by major economic powers, or negotiations in general in which the vulnerable have everything to lose. In short, the judicial process becomes one more victim of neoliberalism, the process having maintained medieval procedural instruments so that, far from the liberal schemes of the 19th century, a court of law is once again a place that is alien to the vast majority of the population.

5. The burden of persuasion: an impossible desire

What about the burden of persuasion? 111 It is little more than a matter of faith. A chimera. An establishment of evidentiary thresholds: probable cause 112, preponderance of evidence 113, clear and convincing evidence 114, and beyond any reasonable doubt 115, which are identical respectively to semi-plena probatio, probatio summaria or probatio prima facie broken down onto two levels, plena probation and probatio plenissima. In short, a simple imitation of old realities in the system of legal evidence that fortunately no longer exist.

Such copying has not really been conscious, and so almost all the old medieval evidentiary standards from which they derive have been forgotten in the Anglo-Saxon sphere. They are not even cited, except for probatio prima facie 116, with exactly the same meaning it had in the Roman-canonical process, i.e., the initial sufficiency of proof, observed at first sight.

The question to be elucidated behind these phrases, and others like them, is whether there can be anything scientific that escapes simple intuition. There undoubtedly was when, as we have seen, plena probatio and semiplena probation were determined precisely. But when we look beyond simply fossilised schemes to achieve epistemic categories, it is extremely difficult, if not impossible. We can attempt to construct these levels on the basis of the plural formulation of hypotheses inherent to probability 117, but this formulation will be more or less complete depending on effort, creativity or even the will of the person doing the work. This implies an element of maximum insecurity that is simply unacceptable. We already know that when it comes to reconstructing past events, we cannot draw conclusions as we would in the world of physics or chemistry because we cannot contrast it with other similar hypotheses. Thus, Bayes’ theorem 118 fails in this context.

This creative formulation of hypotheses is what police, prosecutors and judges have done throughout history. Drawing on their personal and professional experience, they try to put together the pieces at their disposal to create a puzzle that corresponds to that previous experience. They do this mainly by using representativeness heuristic 119, i.e., making a rough statistical calculation that can be quite imprecise or even crude as it simply depends on their intuition governed by the use of the heuristic, in other words, their experience. This has given human beings a sense of justice on many occasions over the millennia. The time has come to find something better without insisting on the initial model or even passing it through the sieve of epistemology.

There are situations where data can completely confirm a fact without the need to go into further complexities. They are more frequent than it would seem, especially in civil proceedings, where documentary evidence is king, not only thanks to written documents, but also the ease with which we can access communications by e-mail, text messages or even recordings from security cameras, and the simplicity of corroborating an alibi with the presence of our electronic devices at a certain place. Such data was inconceivable only 40 years ago, but today we have it, and therefore we can no longer ignore it and continue to rely predominantly on intuition, as the case has been traditionally.

Moreover, scientific evidence, especially biological evidence, has now allowed materials that were once useless to tell us a great deal about the participants in the places where the evidence is found. DNA fingerprints are relatively easy to obtain in a criminal context and can confirm the presence of a person who could only conceivably have been there if they were involved in the event.

We no longer have to rely on the intuitive assessment of the gestures or tone of voice of witnesses 120, and in the vast majority of cases, we can even dismiss these witnesses. Their testimony is always a blurred recollection often prepared by lawyers, removing any credibility, not to mention the intrinsic precariousness of their memory. Without documents or scientific evidence, we could only rely on whether someone had seen something. Today, this is no longer necessary. Documents and expert opinions are the best notaries of reality. Never in history has it been more attainable to judge with the greatest empirical, not intuitive, possibilities of justice.

Consequently, the future of evidence in the process must not depend on precarious standards that are little more than a working method, and which cannot create evidentiary thresholds. On the contrary: the process, and, more specifically, the evidence, must embrace science. The participation of economists, biologists, doctors and psychologists must be much more frequent in our trials, to the point that we cannot do without them, even when it comes to assessing intentionality in criminal proceedings or the basic facts of the subject matter of the trial. Their presence must be formalised, normalised, and not made dependent on the pleadings of the prosecutor or the parties.

After this, the judge will no longer be an omniscient being inspired by divinity, which is what he has tried to be since the times of ancient Egypt 121. They will simply be a manager of the scientific evidence that is accumulating, a guarantor of fundamental rights and an applicator of the law to facts that are determined with increasing certainty by science. Their personal assessment, so often intuitive, will no longer depend on their intimate sense of conviction, to the best of their knowledge, but rather on the ability to correctly compile data from experts with whom they cannot enter into conflict as they do not have the scientific competence to do so.

All this will illuminate a new process, non secundum conscientiam, sed secundum probata peritorum. In the field of evidence, it will only be possible to exercise the defence by discrediting these opinions with other scientific data.

6. Bibliography

Andrews, N., 2013: Andrews on Civil Processes, Cambridge: Intersentia, vol. 1.

Azo, 1567: Brocardica (aurea), rúbrica XX, Basilea

Azo, 1610: Summa Azonis, Venecia

Bar, L, 1867: Recht und Beweis im Zivilprozesse, Leipzig.

Barbosa Moreira, J.C., 1988: “Julgamento e Ônus da prova”, in Temas de direito processual, segunda série, Sao Paulo: Saraiva, p. 75.

Bartolo, 1588: Bartoli a Saxoferrato Commentaria, in II.partem infortiati, Basilea, tít. IV, L. V, p. 468.

Beccaria, C., 1764: Dei delitti e delle pene, reed. de Acquarelli de Bussolengo 1996.

Bentham, J., 1823: Traité des preuves judiciaires, Paris : Bossange Frères, t. II

Besso-Marcheis, 2015: “La vicinanza della prova”, Revista Eletrônica de Direito Procesual, v. 16, p. 93. http://www.e-publicacoes.uerj.br/index.php/redp/article/view/19962/14303

Blackstone, W., 1768, Commentaries on the Laws of England, Lib. I, London: Clarendon Press.

Blackstone, W., 1768, Commentaries on the Laws of England, Lib. III, London: Clarendon Press.

Bracton, H., 1569: DeLegibus et Consuetudinibus Angliae, London, Lib. III, Cap. VIII.

Bruni, F., 1546: Tractatus de indiciis, et tortura, Lyon.

Bulgari, Summa de judiciis, Tit. 53, p. 91, in Wunderlich, A., 1841: Anecdota quae processum civilem spectant, Göttingen.

Calamandrei, P., 1939: “Il giudice e lo storico”, Rivista di diritto processuale civile, XVII, p. 105 .

Chiovenda, G., 1923, Principi di Diritto Processuale, 3ª ed. Napoli: Jovene.

Cohen, L. J., 1977: The probable and the provable, Oxford.

Colish, M. L., 1999: Medieval foundations of the western intellectual tradition, 400–1400, London: Yale University Press, p. 25, 265-266, 319 .

De Fano, M., 1584 : Negativa qualiter probanda, in: “AAVV, Tractatus illustrium in utraque tum Pontificii, tum Caesarei iuris facultate Iurisconsultorum, De Probationibus”, T. IV,Venecia, p. 12, n. 3.

Decoeur, H., 2011 : «Maat, entre Cosmologie et Mythe: Le Principe Constitutionnel d’un Etat de Racine Chtonienne en Ancienne Egypte.» Revue Juridique Themis, vol. 45, n. 2, p. 343 .

Dei Vecchi, D., 2020: Los confines pragmáticos del razonamiento probatorio, Lima: Zela 2020.

Dennis, I.H., 2013: The Law of Evidence, Mytholmroyd: Sweet & Maxwell.

Döhring, E., 1964: La prueba, su práctica y apreciación, Buenos Aires: Ediciones Jurídicas Europa-América.

Duranti, G., 1585: Speculum iuris, p. II, Veneci

Eckhardt, K.A., 1953: Lex Salica, 100 Titel-Text, Weimar: Hermann Böhlaus Nachfolger

Edwards, W. / Lindman, H. / Savage, L. J., 1963: “Bayesian Statistical Inference for Psychological Research”, Psychological Review, 70, p. 193.

Fernández López, M., 2007: “La valoración de las pruebas personales y el estándar de la duda razonable”, Cuadernos Electrónicos de Filosofía del Derecho, núm. 15.

Ferrer Beltran, J., 2019: “La carga dinámica de la prueba. Entre la confusión y lo innecesario”, in Nieva Fenoll, Ferrer Beltrán, Giannini, Contra la carga de la prueba¸ Madrid: Marcial Pons, p. 73.

Ferrer Beltrán, J., 2021: Prueba sin convicción, Madrid: Marcial Pons.

Finkelstein, M. O., 2009: Basic concepts of probability and statistics in the Law, New York: Springer, p. 11.

Gascón Abellán, M., 2005 “Sobre la posibilidad de alcanzar estándares de prueba objetivos”, Doxa, n. 28, p. 127 .

Ginther, M.; Cheng, E.K., 2018: “Surprise vs. Probability as a Metric for Proof”, Seton Hall Law Review 48, n. 4, p. 1081 .

Glaser, J., 1883: Beiträge zur Lehre vom Beweis im Strafprozeß, Leipzig: Duncker & Humblot, p. 85.

Glaser, J., 1883: Handbuch des Strafprozesses, vol. I, Leipzig, Dunker & Humblot.

González Lagier, D., 2020: “¿Es posible formular un estándar de prueba preciso y objetivo? Algunas dudas desde un enfoque argumentativo de la prueba”, Revista telemática de filosofía del derecho, n. 23, p. 79.

Gordon, W., 2007: “A Comparison of the Influence of Roman Law in England and Scotland”, in Roman Law, Scots Law and Legal History: Selected Essays, Edinburgh: Edinburgh University Press.

Hall, C. C., 1840: De indiciis, eurumque vi, ad probationem in causis poenalibus efficiendam, Copenhague.

Hans, V. P. / Eisenberg, T., 2011: “The Predictability of Juries”, DePaul Law Review, 60, p. 375.

Herlitz, G. N., 1994-1995: “The Meaning of the Term Prima Facie”, Louisiana Law Review 55, no. 2, p. 391.

Hossfeld, F., 1665: De indiciorum materia, Altdorf.

Jaumar y Carrera, J., 1840: Práctica forense, Barcelona: Boet y cía.

Kagehiro, D. K. / Stanton, C., 1985: “Legal vs. Quantified Definitions of Standards of Proof”, Law and Human Behavior, vol. 9, n. 2, p. 160.

Kahneman, D. / Tversky, A., 1982: Subjective probability: A judgment of representativeness, in: “Kahneman / Slovic / Tversky (ed.), Judgment under Uncertainty: Heuristics and Biases. Cambridge: Cambridge University Press, p. 33.

Kaser, M. / Hackl, K., 1996: Das römische Zivilprozessrecht, München: C.H. Beck.

Kleinknecht, T. / Meyer, K. / Meyer-Goßner, L., 1995: Strafprozeßordnung, München: C.H. Beck.

Kohler, J. (con Holtzendorff), 1904: Encyclopädie der Rechtswissenschaft, vol. 3, Leipzig: Duncker & Humblot, p. 315.

Lara Peinado, F., 1997: Código de Hammurabi, Madrid: Tecnos.

Laudan, L., 2012: Truth, error and Criminal Law, Cambridge University Press.

Laumen, H-W., 2009, „Grundbegriffe der Beweislast“, in Baumgärtel / Laumen / Prütting, Handbuch der Beweislast, München: Heymanns, p. 27.

Leipold, D., 1997: Comentario al §286 ZPO, in: “Stein/Jonas, Kommentar zur ZPO, Tübingen: Mohr Siebeck, p. 527.

Loftus, E., 1996: Eyewitness testimony, Harvard: Harvard University Press.

Manzanero, A. L., 2008: Psicología del testimonio, Madrid: Pirámide.

Mazzoni, G., 2015: Psicologia della testimonianza, Roma: Carocci, p. 108 .

Mitidiero, D., 2020: O ônus da prova e seus inimigos, Revista de processo, n. 306, p. 17.

Montero Aroca, J., 1981: La justicia municipal, in: “Estudios de Derecho Procesal”, Barcelona: Bosch, p. 91.

Mueller, C. B. / Kirkpatrick, L. C., 2003: Evidence, New York: Wolters Kluwer.

Muñoz Sabaté, Ll., 1967: Técnica probatoria, Barcelona: Praxis.

Muñoz Sabaté, Ll., 1992: Tratado de probática judicial, Barcelona: Bosch.

Muñoz Sabaté, Ll., 2008: Summa de probática civil. Cómo probar los hechos en el proceso civil, Madrid: Wolters Kluwer.

Newton, E.F. (ed.), 1909: Fallos de la Cámara Federal de Apelación de La Plata, t. XV, La Plata, p. 275.

Nieva Fenoll, J. / Ferrer Beltrán, J. / Giannini, L., 2019: Contra la carga de la prueba, Madrid: Marcial Pons.

Nieva fenoll J., 2014: “El mal nombre del principio inquisitivo”, Justicia, n. 1, p. 131.

Nieva fenoll, J. 2014: “la cattiva reputazione del principio inquisitorio”, Rivista Trimestrale di Diritto e Procedura Civile, vol. 68, n. 3, p. 943.

Nieva Fenoll, J. 2017: “La actuación de oficio del juez nacional europeo, Justicia, n. 1, p. 181.

Nieva Fenoll, J. 2020: “Carga de la prueba y estándares de prueba: dos reminiscencias del pasado”, InDret 3, p. 406.

Nieva Fenoll, J., 2018: “La carga de la prueba: una reliquia histórica que debiera ser abolida”, Revista Ítalo-española de Derecho Procesal, n. 1, p. 129.

Nörr, K. W., 2012: Romanisch- kanonisches Prozessrecht, Berlin: Springer, p. 127.

Pardo, M. S., 2018: “Epistemology, psychology, and standards of proof: An essay on risinger’s surprise theory”, Seton Hall Law Review, 48(4), p. 1039.

Passanante, L., 2020: “Per la difesa dell’onere della prova”, in Saccoccio; Cacace (ed.), Europa e America Latina. Due continenti, un solo diritto, Torino y Milano: Giappichelli, p. 798.

Patetta, F., 1890: Le ordalie, Torino: Bocca.

Peters, K., 1985: Strafprozeß, München: C.F. Müller.

Peyrano, J. W. (dir. y coautor), 2008: Cargas probatorias dinámicas, Buenos aires: Rubinzal-Culzoni.

Pothier, R. J., 1827, Traité des obligations, in Ouvres de Pothier, T.I, Paris.

Ramos, V. P., La carga de la prueba en el proceso civil, Madrid 2020.

Redmayne, M., 1999: “Standards of Proof in Civil Litigation”, Modern Law Review 62, n. 2, marzo, p. 172.

Redmayne, M., 1999: “Standards of Proof in Civil Litigation”, Modern Law Review 62, no. 2, Marzo, p. 167 .

Risinger, D., 2018: “Leveraging surprise: What standards of proof imply that we want from jurors, and what we should say to them to get it”, Seton Hall Law Review, 48(4), p. 965.

Robertson, C. B., 2012-2013: “The Right to Appeal”, 91 ,The North Carolina Law Review, p. 1219.

Rosenberg, 1923: L. Die Beweislast, Berlin: Otto Liebman.

Rosenberg, L. / Schwab, K.H. / Gottwald, P., 2011: Zivilprozessrecht, München, Beck.

Russell, B., 1952: “Is There a God?”, in Slater, (ed.). The Collected Papers of Bertrand Russell, Vol. 11: Last Philosophical Testament, 1943–68 Routledge, p. 542.

Sassoon, J., 2001: Ancient Laws And Modern Problems: The Balance Between Justice and a Legal System, London, p. 40.

Serra Domínguez, M., 1991: Comentario al art. 1252 del Código Civil, in: “Comentarios al Código Civil y compilaciones forales” (dir. Albaladejo), Madrid: Edersa, tomo XVI, vol. 2, p. 66.

Struve, G.A., 1666: De indiciis, Jena.

Taruffo, M., 2005: La prueba de los hechos, Madrid: Trotta.

Taruffo, M., 2019: “Casi una introducción”, in Nieva Fenoll, Ferrer Beltrán, Giannini, Contra la carga de la prueba¸ Madrid: Marcial Pons, p. 18-21.

Wach, A., 1872: “Der Entwurf einer deutschen Civilprozeßordnung“, Kritische Vierteljahresschrift, 14, p. 357.

Whitman, J. Q., 2005: The origins of reasonable doubt, New Haven y London: Yale University Press.

Wigmore, J. H., 1904: A Treatise on the System of Evidence in Trials at Common Law, Boston: Little, Brown.