INTRODUCTION

Exceptional evidence is a theme little faced by legal theory. In Brazil, however, they have special practical relevance, either because they are legally provided for in the final part of article 155 of the Code of Criminal Procedure, or because their practical use has led researchers to conclude that the information center of the criminal process has been shifted to the investigative phase².

In effect, the information obtained in the investigation is understood as informative elements in the Brazilian criminal process. In other words, they do not have the nature of evidence because they are not subject to the adversarial process, called by Gomes Filho the "cornerstone" of the discipline of evidence in the criminal process³. Nevertheless, it is possible, even if performed during the investigation, that certain steps be considered exceptional evidence: they are the precautionary evidence, the nonrepeatable evidence and the anticipated evidence.

Strictly speaking, information considered as evidence is apt to be used by the judge to form a condemnatory judgment, while the informative elements of the investigation do not lend themselves to this. It should be noted that in one case or another, illicit evidence⁴ is not allowed, including the fruit of the poisoned tree or simply poisoned evidence, or even, as in Portuguese law, the hypotheses of prohibition of evidence⁵. However, what will be dealt with here is all information that can be valued, which excludes those that, by virtue of the legal system, cannot be the basis for a judicial decision, since truth is not an absolute value⁶.

It is necessary, on the other hand, to make it clear that the purpose of the informative elements is not only to make a judgment on the admission of the accusation or complaint, that is, to assess the empirical support for the accusation⁷, but to assess the requirements for any decision prior to the receipt of the accusation or complaint.

Thus, the informative elements should be the empirical support of any judicial decision rendered in the prosecution phase that goes from the investigation to the act of receiving the indictment. Among these measures are personal precautionary measures (temporary custody, preventive custody, personal precautionary measures other than prison and freedom), real or property precautionary measures (sequestration, seizure, and specialization and registration of mortgage), and exceptional evidence (cautionary evidence, non-repeatable evidence, and anticipated evidence).

The doctrine has been addressing the determination of the degree of evidentiary confirmation necessary for the judge to consider proven the factual statement that leads to conviction, which is called the standard of proof⁸ or the evidentiary standard⁹.

However, there is little literature on the degree of confirmation of factual statements that give empirical support to the judicial decisions handed down in the persecutory phase that goes from the investigation to the receipt of the accusation. In this case, even the expressions standard of proof and evidential standard would be an inconsistency from the point of view of legal nomenclature when dealing with the dogmatic rigor of criminal procedure that should be demanded in the use of the most adequate terms, although it must be admitted that as far as the epistemological aspects are concerned, there would be no reparation to be made.

In the case of a criminal procedural text, the institute will be treated using the expression "informative standard" to designate the criterion of sufficiency of the empirical support that any judicial decision should have, in such a way that after receiving the initial accusatory piece, the empirical support should be based on evidence, while before that, the empirical support will be based on informative elements. In any case, it is possible, coherently with dogmatic rigor, to always call it the informative standard.

Having said this, the main purpose of this paper is to investigate whether and what would be the informative standard necessary to consider that there is empirical support for court decisions issued with the purpose of granting evidentiary injunctions and that is able to eliminate an unreasonable degree of subjectivity in decisions to allow their rational controllability.

To this end, two theoretical and bibliographical approaches will be used using the hypothetical-deductive method. The first is preparatory and conceptual on what precautionary evidence measures are, establishing the distinctions for other types of exceptional evidence, as well as comparing the understandings on the relationship or identity between precautionary evidence measures, measures for obtaining evidence and special investigation techniques. The research methodology will be bibliographic, based on Brazilian criminal procedural dogma on the subject, as well as an analysis of the most relevant legal provisions on the institutes.

The second theoretical approach will focus on the understanding of evidence standards and, unlike the previous core, will not be limited to a dogmatic approach, but will include an epistemological research based on Brazilian and foreign researchers, especially Ferrer Beltrán¹⁰, Larry Laudan¹¹, Michele Taruffo¹², Gustavo Badaró¹³ and Vinícius Vasconcellos¹⁴, to reach a conclusion proposing an informative standard of its own for the types of decisions understood as authorizing evidentiary precautionary measures.

1 EVIDENTIARY PRECAUTIONARY MEASURES: LEGAL AND DOGMATIC ASPECTS

As explained in the introduction, exceptional evidence is understood to include cautionary evidence, non-repeatable evidence, and anticipated evidence. However, this paper will not discuss the three types of evidence, only the cautionary evidence.

The first question to be addressed is whether the final provision of article 155 of the Brazilian Code of Criminal Procedure, in listing the exceptional evidence in kind, is equivalent to the types of evidence listed in paragraph 3 of article 3-C of the same legal diploma, whose effectiveness has been suspended by virtue of the injunction decision issued by Justice Luiz Fux in ADI 6298.

This is because when the judge of guarantees was instituted in Brazilian criminal proceedings, the aforementioned paragraph 3 of article 3-C of the Brazilian Code of Criminal Procedure determined that the investigation records or any other measure that falls under the competence of the judge of guarantees should not be attached to the case records, with the exception of unrepeatable evidence, measures for obtaining evidence, and/or anticipation of evidence.

In this sense, it is not difficult to relate the expressions in this paragraph to the final part of the same article 155 of the Brazilian Code of Criminal Procedure. Thus, the "unrepeatable evidence" of § 3 of Article 3-C is the "non-repeatable evidence" of art. 155, and the "anticipation of evidence" of § 3-C is the "anticipated evidence" of art. 155. However, the question is: are the "measures for obtaining evidence" of §3 of article 3-C the "precautionary evidence" of art. 155?

Many authors consider it to be the same¹⁵, however, the understanding presented by Santoro and Gonçalves seems more adequate, in the sense that one should not confuse cautionary evidence with a measure or means for obtaining evidence, nor with the socalled special investigation techniques¹⁶.

Means or measures for obtaining evidence are extra-procedural evidential activities, based on surprise¹⁷, also called hidden methods¹⁸, without the participation of the defense and without direct adversary proceedings, which may serve to obtain sources or elements of proof, which can be classified into four groups: "a) examinations, inspections, searches; b) searches, seizures, kidnappings; c) interceptions, wiretaps, breaches of secrecy; d) special actions for the investigation of organized crime"¹⁹.

Special Investigative Techniques are contemplated in the Palermo and Merida Conventions and include "covert police activity, of a confidential or even secret nature, which is carried out with the purpose of obtaining intelligence flows concerning the activities of suspected persons and/or the gathering of evidentiary material"²⁰.

The fact is that the Special Investigative Techniques are not evidence, but techniques capable of generating evidence, in such a way that if there is no express legal provision that the techniques are considered measures for obtaining evidence, the investigative procedures cannot be considered evidence²¹. In the words of Santoro and Gonçalves, "there is an intrinsic relationship, the power of the SITs to become means of obtaining evidence"²² , but not to be confused with them.

Precautionary evidence must be understood in the scope of precautionary measures. In criminal proceedings there is no precautionary process, but precautionary measures, which can be classified as personal, evidentiary or property²³. Precautionary measures can be determined during the investigation or during the trial. In the first case they are called preparatory and in the second they are called incidental. In all cases, precautionary measures must have their own characteristics and requirements. The characteristics of precautionary measures are jurisdictionality, exceptionality, instrumentality, provisionality, legality, strictness, and contradictory nature.

The requirements for precautionary measures in general are presented by traditional doctrine as fumus boni juris and periculum in mora²⁴.

Critically, its common to object to the use of the expression fumus boni juris in criminal procedure and argue for its replacement by fumus commissi delicti, in fact more appropriate, since it is understood as proof of the crime and evidence of authorship or participation. It would be, to say the least, inappropriate to call a crime (or the empirical support that allows one to say there was a crime and evidence of authorship) "fumus boni juris".

The expression periculum in mora is also challenged, proposing its replacement by periculum libertatis, since the problem in criminal proceedings is not the delay but the risk that the freedom of the accused may circumstantially cause to the process, either to its final result or to the production of evidence. However, this is an objection that, despite being pertinent when dealing with personal precautionary measures, is not exactly pertinent to evidence precautionary measures, whose risk of loss can be effectively provoked by the delay, which would justify the immediate production, before the moment in which the judicial contradictory process takes place, as well as by the previous knowledge of the accused, who could oppose or resist its execution.

The proposal by Santoro and Gonçalves²⁵, adopted here, is to understand that the requirements for any evidentiary precautionary measure are the "need for the evidentiary measure" (instead of fumus boni juris or fumus commissi delicti) and "danger of loss of evidence" (instead of periculum in mora or periculum libertatis).

The "necessity of the evidence measure" would be constituted by (1) the certainty of the crime (which we will call here materiality), (2) the probability (which we will call indications) of authorship or participation, and (3) the pertinence of the measure in relation to the intended purpose. The "danger of loss of evidence" would be constituted alternatively by (1) the risk of loss of the source of evidence due to the lapse of time or (2) the risk of loss of the source of evidence due to the prior knowledge of the investigated.

It has already been established that the Special Investigative Techniques, in order to become means of obtaining evidence, require legal regulation. Now it is necessary to examine whether the measures or means for obtaining evidence are precautionary measures of proof. To do so, it is necessary to check whether the means of obtaining evidence meet all the characteristics and requirements required to be considered an evidentiary precautionary measure.

However, not every evidence obtaining measure meets the characteristics and requirements. The example given by Santoro and Gonçalves is that of the controlled action, which is a means of obtaining evidence regulated by Law 12,850/2013 (therefore, it is a Special Investigation Technique converted into a means of obtaining evidence) that does not require judicial authorization and, therefore, does not meet the characteristic of jurisdiction, essential for its configuration as a precautionary evidence measure. In such cases, the authors conclude,

Means of obtaining evidence that do not meet the conditions to be considered precautionary evidence are not useless, but they cannot give rise to an element of evidence that can be valued by the judge, they are only capable of obtaining sources of evidence, from which elements can be extracted by the subsequent use of the means of proof appropriate to the source obtained²⁶.

It is necessary, on the other hand, to admit that the referred authors were referring to the impossibility of valuing these elements as evidence, but what is being discussed here is not the possibility of valuing these elements for a judgment of conviction, but for a judgment of deferment of evidentiary precautionary measures.

In other words, means of obtaining evidence that do not fulfill all the characteristics of evidentiary precautionary measures cannot be considered apt to lead to a guilty verdict, but may provide empirical support for decisions that determine the execution of evidentiary precautionary measures, although not to be confused with them.

Having thus established the distinctions between the institutes addressed, and having defined the characteristics and requirements of evidentiary precautionary measures, as well as the information that can serve as empirical support for the determination of evidentiary precautionary measures, it is necessary to address the sufficiency criteria, that is, the informative standards for the granting of these measures.

2 STANDARD OF PROOF: CONCEPTUAL AND FUNDAMENTAL ASPECTS

As Ferrer Beltrán puts it, "in the Roman-Germanic and common law traditions, two standards of proof (if they deserve to be called that) are commonly used in criminal procedure: the intimate conviction and the 'beyond reasonable doubt'"²⁷.

Michele Taruffo states that the theme has not received particularly important doctrinal analysis, since the prevailing trend in the literature is to consider the principle of the judge's free persuasion or the method of prudent assessment of evidence. The fact is that although free persuasion is used to exclude the use of legal rules of evidence, such as the one that establishes that a confessed fact has been proven²⁸, for example, it does not point to a significant criterion on how the judge should evaluate the evidence, consolidating understandings in the sense that the judge is free to base his decisions according to his intimate conviction or according to his moral certainty²⁹.

However, if one intends to work with a rational model of evidence, one can never admit that the purpose of the evidential activity is to produce a mental state in the judge, such as conviction, belief or something else. These mental states are not voluntary and, therefore, are not subject to rational justification, the most that could be done would be an empirical psychological investigation of the causes that led the judge to believe in a proposition.

In this sense, it is fallacious to try to distinguish free conviction from intimate conviction based on the requirement of motivation, since it does not seem reasonable to demand that the internal psychological processes that led to a certain conviction be expressed, even because they are little or not at all objectifiable³⁰.

The free evaluation of evidence, devoid of epistemic methods of knowledge, are in reality the expression of intimate conviction, and therefore uncontrollable from a rational point of view.

In addition to the inferential reasoning that the judge must develop when evaluating the information at his disposal, it is necessary to establish a criterion capable of considering that a certain amount of evidence is sufficient to establish that a fact is proven. This criterion is called the standard of proof³¹.

Badaró defines that "standards of proof are criteria that establish the degree of evidential confirmation necessary for the judge to consider a factual statement as proven"³².

For Vinícius Vasconcellos, the Brazilian criminal procedure system requires a rigorous standard of proof to reinforce the presumption of innocence and, although it can be improved from the current legislation, "it is essential to make a legislative change to insert a provision in an express manner and properly delimit its content."³³

However, although a legislative provision is highly desirable, it is important to pay attention to the distinction that Janaina Matida and Antonio Vieira make between rules and standards. "Rules are strategies designed for situations in which the intention is to free agents from the burden of reflection on future decisions", while "the standard represents a different type of strategy. By making use of vague terms in normative formulation, the architects of the most diverse legal systems aim to guarantee space for the discretion of the enforcers.

Thus, the authors give the example of the provision that limits the speed of cars to 80 km/h as a rule and the determination that drivers must drive prudently as a standard. Both have the same purpose, but "the standard is intended to be a fairer rule for concrete cases, because it always reserves to the agent discretion to analyze the individual case and its specifics having the purpose for which it was created."³⁴

But under no circumstances should standard of proof be confused with adherence to judicial subjectivism, since this would be characterized as a model of legal irrationalism. It is necessary to understand the standards as the criterion of evidential sufficiency capable of controlling the rationality of the decision, an intersubjective control. For this it is necessary to define the standard.

This definition is based on political criteria (according to Ferrer Beltrán, criminal policy ³⁵) or axiological criteria³⁶ or both³⁷. The fact is that the standards are designed to admit which errors are more acceptable. Therefore, in a justice system based on the presumption of innocence, a high standard should be used for the confirmation of the accusatory hypothesis, since it is more acceptable that errors occur in order to acquit the guilty than to convict the innocent.

The most commonly used standard has its origin in Common Law and is gaining important space in countries of Roman-Germanic tradition, especially in Brazil. It is the "Beyound a Resounable Doubt" standard, known as BARD.

There are many criticisms of this standard, which we will address below, but it is important to draw attention to the argument put forward by Vinícius Vasconcellos in favor of the BARD. For him, this is a standard created for the jury trial system, in which jurors do not justify their decision. Since in Brazil there is a requirement to provide reasons for judicial decisions, this model may result in a more guaranteebased mechanism than in common law systems³⁸.

Although he admits that the current construction of the BARD may result in damage to the Brazilian system "due to its imprecision and openness"³⁹, Vasconcellos believes it is possible to guide it towards parameters of rational assessment of evidence, not with its complete objectification, because "there will always be room for subjectivism"⁴⁰, but seeking to reduce it as much as possible to increase the controllability of the decision.

Vasconcellos defines “...reasonable doubt as the alternative hypothesis to the incriminating thesis, which proves to be logically possible and supported by the evidence in the process”⁴¹. To this end, he sustains the need to adopt some parameters:

From the partial conclusion that the standard of "proof beyond reasonable doubt" should be adopted in the Brazilian legal system, it is argued that some parameters must be met for its normative provision. Essentially, it is important to emphasize that its definition should occur in two moments.

First of all (1), in order to avoid the criticism on the inversion of the burden of proof, it is necessary to rule that the accusatory party (as a rule, the Public Prosecution Service) must prove all the elements of its incriminating hypothesis in a consistent manner, based on evidence lawfully produced in adversary proceedings. The accusatory hypothesis must be able to coherently and fully explain all the factual elements proven in the process in an individual and specific way, presenting available confirmatory criteria. According to^{Badaró (2019}, p. 255), "for a standard of proof to be complete, it must require evidence that supports all the facts alleged by the prosecution and that are criminally and procedurally relevant."

Then (2) - and here is the fundamental difference with the civil standard of "preponderance of evidence" - after consistently proving the incriminating hypothesis, possible alternative explanations for the proven facts must be ruled out, i.e., the incriminating thesis must resist any reasonable doubt. This is a falsifiability check of the incriminating hypothesis (SCHIAVO, 2013, p. 91), in which the plausibility of the alternative hypotheses must be analyzed (CATALANO, 2016, p. 90; CANZIO, 2004, p. 304)⁴².

However, both in Brazil and abroad, jurists and epistemologists are critical of BARD for its conceptual imprecision and subjectivity.

3 MAIN CRITICISMS TO BARD AND ALTERNATIVE PROPOSALS: THE PROBLEM OF SUBJECTIVATION OF IN DUBIO PRO REO

Larry Laudan clarifies that in a creative decision⁴³, the Supreme Court understood that the United States Constitution “required instructing all criminal jurors about what BARD was a minimum threshold to convict”. The more traditional instruction was that proof beyond a reasonable doubt meant “belief with moral certainty”. However, since Victor v. Nebraska, the moral certainty formula has been rejected, as it could imply a legitimization of a moral or emotional judgment, not based on the evidence.

Moving beyond the more traditional definition of moral certainty, Laudan presents what he understands to be the five alternative explanations that do not exhaust attempts at explanation: (1) BARD as the certainty in the belief appropriate for important decisions in life, such that jurors were directed to decide in the same way they make major decisions in their lives when they are certain of the belief on which they rely; (2) reasonable doubt as the kind of doubt that would make a prudent person waver about whether or not to act based on it; (3) BARD as the belief in guilt beyond all reasonable doubt when the juror has an enduring conviction that the accused is guilty; (4) reasonable doubt is one for which a reason can be given; (5) BARD as a high probability⁴⁴.

It seems clear that all attempts at explanation run into a clear conceptual inaccuracy, either because they are based on the exclusive subjectivity of the judge, as in the case of the first two; or because they try to replace the subjectivity with an absolutely imprecise temporal formula, as in the third; or because they create a sophism, using the definition as an explanation (what cannot be given reason is unreasonable) going around in circles; or because they replace an imprecise formula (reasonable doubt) with another formula of similar imprecision (high probability).

In view of the criticism of the subjectivity of the standard of proof most commonly used in Western countries, attempts are not few to present a standard that accounts for the necessary compatibility with the political decisions that guide the legal system, especially with the presumption of innocence.

For Ferrer Beltrán, a standard of proof must meet three requirements: (1) the degree of corroboration of the hypothesis cannot depend on the beliefs of the decisionmaker; (2) the formulation must be such that intersubjective control of its application becomes possible; and (3) a political aspect must be added, consisting in the decision regarding preference over possible errors (for example, giving preference to false acquittals over false convictions). Thus, he presents a proposal for an evidentiary standard for the accusatorial hypothesis:

1) The hypothesis must have a high level of contradictoriness, explain the available data, and be able to predict new data that has in turn been corroborated.

2) All other plausible hypotheses explaining the same data, which are compatible with innocence, must have been refuted⁴⁵.

Ferrer Beltrán uses the expression "contrastación", which is translated as contradictoriness. The author, when dealing in another work with how a hypothesis can be corroborated or contradicted, explains that the first act of the moments of substantial importance in the evidential process is what he called discovery, which begins with the generation of the hypothesis, proceeds to the elimination of the hypothesis and to the structuring of the argumentation. For the hypothesis to be consistent, it must meet the requirements for its formulation: (1) it must be well formed, logically consistent and significant; (2) it must be based on existing knowledge; (3) it must be empirically contradictory. From a legal point of view, two more requirements are required: (1) this contradictory nature must be not only potential, but also real; (2) it must have legally relevant objects. The degree of contradictory nature of the hypothesis is a function of (a) the predictions that have turned out to be true that could be made from the available data and (b) the improbabilities that another hypothesis could account for the same data and make the same true predictions⁴⁶.

Therefore, the first element of the standard proposed by Ferrer Beltrán is linked to the ability of the hypothesis to explain the available data and predict new data also corroborated.

The second element is centered on the refutation of alternative explanatory hypotheses that are compatible with innocence. Ferrer Beltrán, however, rules out three possibilities: (1) hypotheses that are not plausible; (2) hypotheses that are not compatible with the data available in the case; and (3) ad hoc hypotheses.

The author considers non-plausible those that are not compatible with the current state of knowledge or that have no basis in knowledge to support themselves. His example is of someone who, to justify that stolen jewelry was found in his pockets, claims that he was abducted by Martians who put the jewelry there.

As an ad hoc hypothesis, Ferrer Beltrán indicates hypotheses constructed a posteriori that can be used to be used in any case, such as, for example, the hypothesis of a plot against the accused.

Finally, Ferrer Beltrán clarifies that this demanding standard of proof for the accusatorial hypothesis does not apply to the defense hypothesis, "so the standard for the latter does not have to be so demanding.

Badaró criticizes BARD by stating that it is a standard that is supported by a non-rationalistic conception of the purpose of evidence that is intended to generate a belief in the judge. He understands that the assessment serves to verify whether or not the factual hypotheses are confirmed by the evidence, therefore what should be assessed is the degree of confirmation that the standard requires for the hypothesis to be considered proven.

However, Badaró, while acknowledging the merit of considering the relationship between evidence and factual statement, and not between evidence and the judge's conviction, also criticizes Beltrán's proposal. For him, the first part errs by requiring the prediction of new facts, while the second part does not make sense to require that, in addition to the defensive hypotheses, the judge should look for ad hoc hypotheses.

Thus, Badaró starts from two premises: (1) the only hypothesis to be proved is the accusatory hypothesis and (2) the accusatory hypothesis must be proved in relation to each of the factual segments that compose it, and not as an inseparable whole, to propose that

the standard of proof in criminal procedure, for there to be a conviction, must be: a) there is evidence that confirms, with very high probability, all the factual propositions that make up the accusation formulated by the prosecution; and, b) there is no evidence that makes it feasible that a fact other than any factual proposition that makes up the accusation has occurred⁴⁷.

There is no doubt that Beltrán and Badaró's proposals are objectionable, that is, they are intended to stand out from the subjectivity of the other commonly adopted standards, be it the intimate conviction or even the BARD.

However, Badaró is right when he says that Beltrán works with the idea of prediction of new facts, when in reality the corroboration of the accusatory hypothesis, which is about the verification of propositions of facts that have already happened, in the past, is enough.

An additional advantage of Badaró's proposal is that it considers an undeniable reality: the accusation does not formulate a factual hypothesis; in reality, the accusation presents several factual propositions and all of them must be confirmed by the existing data. However, the introduction of a superlative ("with very high probability") ends up allowing a subjective graduation to be established, despite the fact that it is mandatory to recognize that there is no psychological element (such as belief, conviction or conviction) in the definition.

It does not seem, however, that any alternative explanatory hypothesis that has not been presented by the defense should not be considered by the judge, even because the defense may not present any explanatory hypothesis that, even so, the presumption of innocence rule should serve as a political foundation for the construction of a standard of proof for the criminal process.

It is, therefore, necessary to think of a standard of proof that does not conceptually depend on the defense's performance to rule out the accusatory hypothesis, even because this is the only one that needs to be presented and proven.

The fact is that “the standard of proof, whatever it may be, is not incompatible with the rule of free persuasion”⁴⁸. This firm affirmation by Badaró should be understood as a kind of warning to the inexistence of a bond between the rule of free persuasion and the rule of presumption of innocence, which is a parameter resulting from a political decision, axiologically oriented, for the establishment of a standard of proof for trial.

The presumption of innocence rule implies the distribution of the burden of proof on the prosecution and establishes a model of uncertainty to base the judgment rule: the in dubio pro reo.

The problem is that although in dubio pro reo is presented in theory as a guarantee of a judgment that makes a political choice by admitting more errors consisting of acquitting the guilty than convicting the innocent, in practice it does not have the expected performance because it is based on concepts strictly linked to subjectivity: the doubt or the certainty of the judge.

In effect, doubt and certainty are states internal to the individual and have much more to do with intimate conviction than with objective criteria for accepting a hypothesis as proven. It is possible that the judge has no proof, but is sure about the facts, whether it is based on his religious beliefs, his life experience or even his professional experience. In dubio pro reo, therefore, is a formula that is based on subjectivity⁴⁹ and therefore cannot be taken as a standard of proof⁵⁰, nor is it possible to think of a standard based on it. It is necessary to think of an objective⁵¹ proposition⁵² as a rule of judgment consistent with the presumption of innocence⁵³ in the following terms: if the necessary degree of evidentiary confirmation is not obtained, the decision must be in favor of the defendant.

Once it is understood that the degree of confirmation can be established objectively, belief is removed as a determining factor of the "doubt" element that underlies the judgement rule in dubio pro reo.

4 THE NEED TO ELABORATE SEVERAL OBJECTIFIABLE STANDARDS FOR DECISIONS OF DIFFERENT NATURES

It is necessary, then, to establish what the necessary degree of confirmation would be, that is, what the standard of proof would be. At this point, it is pertinent to introduce a new element: there is not only one informative standard in criminal procedure. In other words, there is not just one criterion that establishes the degree of evidential confirmation necessary for the judge to consider a factual statement as proven.

This is because, as Badaró well puts it, "[u]nce epistemology provides a range of options of varied standards of proof that can be adopted in different types of proceedings, or even in different stages of the same proceeding, it will be up to the law to define which will be this model of ascertainment”⁵⁴.

There are several phases in the criminal process, which require different standards to be set. Moreover, not only should the standards be different according to the stages of the process, but also according to the decision that will be rendered. In this vein, Beltrán explains that “la dinamica del proceso penal parece exigir diversas stándares de prueba para distintos tipos de decisiones”⁵⁵.

Therefore, if it is true that at the end of the process a condemnatory decision requires a very strict standard, it is no less true that throughout the criminal process extremely invasive decisions are made, such as those that cautiously remove the right to liberty (decree of preventive detention) or that cautiously remove fundamental rights for the production of evidence (telephone interception). It is essential that standards for these decisions be defined, all guided by the political option of the constituent legislator for the presumption of innocence.

Furthermore, as previously mentioned, it should be recognized that some judicial decisions need to have as empirical support that which the criminal procedure legislation and the Brazilian dogma do not consider evidence, but informative elements and, in this way, it will be better to treat from this point on the criteria of sufficiency of factual demonstration with empirical basis of informative standards and not standards of proof.

Therefore, it is necessary to build informative standards for the criminal procedure that meet the following requirements: (1) there should be different standards for the different types of decisions to be made in the criminal procedure; (2) informative standards should be objective, never dependent on the judge's mental states, such as belief, conviction, certainty, doubt, etc.(3) in a criminal trial, the prosecution does not formulate a single factual hypothesis, but an imputation that is based on the allegation of the occurrence of a complex of factual hypotheses, all of which must be considered sufficiently demonstrated; (4) the standard is a criterion of sufficiency of the empirical support of the facts alleged by the party who has the burden of proof, therefore the prosecution, which is why the concept must consider its imputations and not the eventual (and not mandatory) defensive allegations; (5) what needs to be demonstrated are factual imputations, that is, allegations of occurrences of factual hypotheses and not historical narratives; (6) it is desirable that any definition that intends to exclude subjectivity should not use formulations that depend on evaluations that may vary according to the judge, such as “reasonable”, “extremely high”, “plausible”, not including the modal adverbs that denote degrees of identification because they have a known current meaning, such as “probably” and “possibly”.

In light of what is proposed here, it is fundamental that at least six (6) informative standards be defined according to the nature of the decision to be rendered and all must meet the six requirements presented above: (1) standard for conviction; (2) standard for the receipt of the accusation or complaint; (3) standard for the decree of provisional arrest⁵⁶; (4) standard for the decree of non-custodial personal precautionary measures or real or property precautionary measures; (5) standard for the granting and extension of evidentiary precautions and (6) standard of proof for the indictment.

The empirical basis for each of these informative standards varies according to the decision. Thus, for conviction and indictment, only evidence may be evaluated. For the indictment, information from the investigation or exceptional evidence is admissible. For the decree of provisional arrest, elements of information from the investigation are allowed, if the decision is made before or at the time of the accusation, but after this procedural moment, only evidence is allowed. For the granting of precautionary measures to protect property or evidence, information from the investigation may be used if the decision is made during the investigation until the indictment is received, but after that it must be based on evidence.

It must be admitted that what has been stated above would have clearer and more objective legal support if §3 of art. 3-C of the Brazilian Code of Criminal Procedure were effective. In any case, it is understood that the interpretation presented here is the only one consistent with a criminal procedure founded on human and fundamental rights.

In order to fulfill the objective proposed in the introduction of this work, it is necessary to approach two informative standards: the standard of proof for guilty verdict (which can properly be called a standard of proof because it is based only on evidence) and the informative standard for the granting of evidentiary precautionary measures.

5 FINAL CONSIDERATIONS

In Brazilian criminal procedure, exceptional evidence is understood to be that which, despite being produced in the phase prior to the filing of the criminal action, may be used by the judge to assess whether it is capable of empirically substantiating a guilty verdict. Precautionary evidence, non-repeatable evidence, and anticipated evidence are considered exceptional evidence.

Precautionary evidence is not to be confused with Special Investigative Techniques, nor with the means of obtaining evidence. Precautionary evidence is a type of precautionary measure that in criminal proceedings can be personal, property or evidentiary. To be considered an evidentiary precautionary measure it is necessary that the following characteristics of a precautionary measure are present: jurisdictionality, exceptionality, instrumentality, provisionality, legality, strictness and contradictory nature.

Furthermore, it is necessary that the requirements for its granting are present. The requirements for provisional evidence measures are understood to be the "need for the evidence measure" and the "danger of loss of evidence".

The "necessity of the evidence measure" would be constituted by (1) the materiality of the crime, (2) evidence of authorship or participation, and (3) the pertinence of the measure in relation to the intended purpose. The "danger of loss of evidence" would be constituted alternatively by (1) the risk of loss of the source of evidence due to the lapse of time or (2) the risk of loss of the source of evidence due to the prior knowledge of the investigated.

For any judicial decision to be considered empirically based, it must be supported by data. Some decisions can only be based on data that is considered evidence, other decisions may be supported by information obtained during the investigation. An example of the first case is the guilty verdict, which can only be supported by data obtained from evidence, whether produced during the trial under judicial contradiction or exceptional evidence. Examples of the second case are the decision to accept the accusation or complaint, which may be empirically based on informative elements or on exceptional evidence, such as the decisions that determine the execution of evidentiary precautionary measures, especially the preparatory ones, that is, those that were issued before the filing of the criminal action.

In any case, it is necessary to establish a criterion of evidential sufficiency to determine that the alleged factual statements may be considered proven, which is called the standard of proof. The important point is that the standard required for each decision is different, since it must vary according to the phase of prosecution. For decisions to be handed down in the prosecution phase, from the investigation to the receiving of the accusation, it is not pertinent to call this criterion of sufficiency of evidence standard, since it is not technically supported by evidence, therefore we call it informative standards.

Thus, it is possible to point out the need to define at least 6 (six) informative standards, to be defined according to the nature of the decision to be made and all must meet the six requirements previously presented: (1) standard for conviction; (2) standard for the receipt of the accusation or complaint; (3) standard for the decree of provisional arrest; (4) standard for the decree of personal non-prison precautionary measures or of real or property precautionary measures; (5) standard for the granting and extension of evidentiary precautions and (6) standard of proof for the indictment.

The definition of an informative standard should be consistent with the presumption of innocence and capable of allowing intersubjective controllability of the decision.

The proposed informative standard for granting an evidentiary injunction in criminal proceedings that is formulated here is the following: the complex of factual allegations presented by the plaintiff is the most probable according to the data available to the judge.

Thus, it is up to the petitioner of the provisional evidence measure to provide information that demonstrates that there is (1) materiality of the crime; (2) evidence of authorship or participation; (3) pertinence of the measure in relation to the intended purpose (therefore, it is up to the petitioner to present the purpose); and, alternatively, (4) the risk of loss of evidence due to the lapse of time or (5) risk of loss of evidence due to prior knowledge of the investigated. The first 3 (three) comprise the "need for the evidentiary measure" and the last two, alternative, are the "danger of loss of evidence", which constitute the requirements of the provisional evidentiary measures.

This proposed standard uses the expression "data available to the judge" in such a way that it is not restricted to evidence, but also encompasses the informative elements of the investigation. It does not handle psychological processes of doubt or certainty, allowing for intersubjective control. It does not require that the applicant's factual allegations be the only explanatory hypothesis in the face of the informative elements, since the purpose of the evidentiary precautionary measure is exactly to rule out the possibility of another explanatory hypothesis with the production of information. Finally, it establishes the need for the factual propositions to be the most probable, and not only possible, in the face of the existing informative elements in order to preserve the fundamental rights threatened by an evidentiary precautionary measure, in coherence with the presumption of innocence and a criminal process founded on respect for human rights.