INTRODUCTION
In the procedural and theoretical literature concerning the notion of proof, it is common to find the use of certain notions, such as belief, certainty, and conviction of the judge in relation to the facts of the case. The emphasis of discussion of these notions is on the states of mind involved in the judge's or jury's determining of the facts. Indeed, on many occasions, it is claimed that the aim of proof is precisely this: that the decision-making body reaches one of these states of mind (i.e., conviction, certainty, absence of doubt, etc.) with respect to the factual premises being proved.1
I should point out that what follows is, or tries to be, a work of general theory of law. For that reason, the analysis tries to be independent of any specific legal regulation of any given legal order. In short, my analysis attempts only to take into account the existence of an organ, judge, or jury that is making decisions that, with or without the constraints of express motivation, involve the proof of factual premises. Furthermore, it may appear that the bibliography and some of the examples tend to be somewhat biased toward civil law, but this is, I hope, purely coincidental.
A good introduction to this can be found in M. Valdés, Introducción, in PENSAMIENTO Y LENGUAJE. PROBLEMAS EN LA ATRIBUCIÓN DE ACTITUDES PROPOSICIONALES 5–22 (M. Valdés ed., 1996).
For this reason, it is interesting to analyze the whole result of the proof activity (using all the information brought to the proceedings) in terms of the propositional attitude of the trier to those proposed facts that are to be proven. What is, however, the propositional attitude involved in utterances which declare facts as proven? Without going into it exhaustively, I believe three distinct answers to this question stand out. First, we can link the proof of a proposition to whether or not the trier believes in the truth of that proposition (this is, in fact, with a variety of formulations and slight alterations, the most widely held thesis in European procedural practice). A second, more hard-line thesis links the proof of a proposition to knowledge. Thus saying that a proposition p is proven is the same as saying that p is known. Finally, the third possible alternative links proof of a proposition to the acceptance of the truth of that proposition. An analysis of these three theoretical options follows. But the discourse should be taken one step further, and the proof of p should be explained in terms of its acceptability (and not simply of its acceptance). Acceptability is a notion that belongs to justifying discourse, not to explanatory discourse. I shall conclude this work with the idea that P is proven if there is sufficient evidence in its favor, that is, if it is acceptable as true.
PROOF AND PROPOSITIONAL ATTITUDES
“It Is Proven That p” as Related to a Fact Finder's Belief in p
Many authors over the years have talked about judicial declarations of proven facts in terms of the beliefs3
It is simply not possible to review even briefly the debate about the concept of belief in contemporary philosophy. In fact, this concept has been the preferred object of philosophical analysis on the part of the so-called philosophy of mind. See the introductions by L. VILLORO, CREER, SABER, CONOCER 11–37 (1982); D. QUESADA, SABER, OPINIÓN Y CIENCIA 35–43 (1998); and S. Saab, Creencia, in 20 EL CONOCIMIENTO, ENCICLOPEDIA IBEROAMERICANA DE FILOSOFÍA 63–87 (L. Villoro ed., 1999). One of the ways of defining “belief” that has won most support in this debate is the one characterized by two elements, that is, (i) the representation of p, and (ii) the disposition to act as if p were true; see R. B. Braithwaite, The Nature of Believing, PROC. ARISTOTELIAN SOC'Y 33 (1932–1933), reprinted in KNOWLEDGE AND BELIEF 30 (A. P. Griffiths ed., 1967).
In the field of theory of law we can mention, among others, the theses defended by two authors—Frank (1951) and Celano (1995)—who hold somewhat separate views in other aspects. Thus Frank says:
Most suits, then, are “facts suits”—suits in which the decisions depend solely on the beliefs of the trial judges or juries about the past events the occurrence of which is in dispute. Prediction of future decisions in “fact suits” not yet commenced means prediction of those future beliefs of trial judges or juries about the past facts. To predict such a decision, one must prophesy that the trial judge or jury will or will not believe that Ding was driving at 80 miles an hour when he hit Dong, or that Nervous did shoot and kill High, or that Simple did hand the deed to his farm to Simon. (Frank 1951, 555)
This is also the majority view in the field of procedural-law scholarship in a good many civil-law countries.4
Such is the case, for example, among Spanish experts in procedural law who maintain that the main objective of the trial is “to convince the judge of the veracity of facts which are claimed to exist in reality”; V. CORTÉS DOMÍNGUEZ, V. GIMENO SENDRA & V. MORENO CATENA, DERECHO PROCESAL CIVIL. PARTE GENERAL 231 (3rd ed. 2000). In a similar vein see, e.g., J. GUASP,1 DERECHO PROCESAL CIVIL 300–301 (1956); and M. MIRANDA ESTRAMPES, LA MÍNIMA ACTIVIDAD PROBATORIA EN EL PROCESO PENAL 45–47 (1997), which include plenty of additional bibliographical references. Finally, this concept has been incorporated into Spanish positive law on occasion; thus, according to Article 741 of the Criminal Justice Law (Ley de enjuiciamiento criminal; hereafter the LECr) “The Court, passing judgement according to conscience on the evidence provided during the trial, the statements made by the prosecution and the defense and those statements made by the accuseds, shall pronounce sentence within the terms fixed by this Law.” (emphasis added) This article has been interpreted for decades by the Spanish Supreme Court as being of a subjective and irrational nature. The Sentencia del Tribunal Supremo (Supreme Court decision; STS) of December 22, 1980, is a good example of this: “article 741 of the LECr attributes to the Court an absolute and sovereign faculty to evaluate the evidence, in which there is neither hierarchy nor pre-eminence of one means of proof over another, and by means of which, this Court may freely form its conviction with respect to the actual problems of fact in question referred to in the trial, without being subject to rational criteria of reasoned judgment, logic, or anything else that is not its sound and impartial conscience.” This and other examples, as well as a proposed interpretation of Article 741 of LECr which tries to make it compatible with the judicial obligation to provide reasoned verdicts, can be found in IGARTUA, 1995, 34–36. More recently, the jurisprudence of the Spanish Constitutional Court (since publication of STC 31 from July 28, 1981) has been introducing limitations to the irrationalist interpretation of Article 741 of the LECr. This line of jurisprudence has also been accepted, although not to its fullest extreme, by the Supreme Court, which now says, for example, that: “the mere subjective certainty of the Criminal Court that there has indeed been an effective demonstration of proof of the charges from which the guilt of the accused is deduced is not enough. Judgement according to “conscience” to which this legal precept refers need not be understood as, or become equivalent to, the closed-off, unshakable, personal and intimate criteria of the fact finder, but rather as a logical appreciation of the proceedings, not without certain guidelines or directives of an objective nature, which results in a re-telling of the facts in concordance with all the relevant supporting data, great or small, that it has been possible to gather together in the proceedings… “Rational criteria” goes hand in hand with logic, science and experience, leaving behind arbitrariness, supposition and conjecture” (STS, November 12, 1996).
the word “proof” identifies…the judge's psychological state of conviction concerning the veracity of all or some of the facts alleged by the parties. In effect, a statement of fact will not be “proven”…if finally…it does not excite in the judge the certitude of the physical reality of the event described in that statement. (Cabañas 1992, 21)
As the quotation from Cabañas shows, the notion of belief is sometimes referred to using the terms “certitude” or “conviction.” In this context I do not find any relevant difference in the meanings of these terms, and therefore it seems fair to treat them as synonyms. On other occasions, on the other hand, proof is linked to certainty relating to the facts. The notion of certainty moves back and forth in the legal doctrine, being for some clearly a notion of belief (Devis Echandía 1981, 34; Naylor 1985, 428) while for others it is a notion of knowledge (Hélie 1866–1867, 323).
Finally, in common-law systems, it is also common to find references to states of mind that relate to the notion of belief—although other names may be used. Thus the standard of proof required for a guilty verdict in a criminal trial uses the formula “beyond a reasonable doubt.”5
This is not the time to analyze in detail how standards of proof work. Suffice it to say that a standard of proof is no more than a rule for decision-making that indicates the minimum amount of corroboration required for a hypothesis to be considered proven. For this reason, if the standard refers to some kind of belief held by a person, i.e., the trier, then the result as to whether the hypothesis is proven or not will depend on that belief being held; L. Laudan, Is Reasonable Doubt Reasonable? 9 LEGAL THEORY 318 (2003); L. LAUDAN, TRUTH, ERROR, AND CRIMINAL LAW, 51–54 (2006).
For more on this and a critical analysis of the rationale of this standard as a decision-making rule, see Laudan, Is Reasonable Doubt Reasonable? supra note 5, at 299–312; L. LAUDAN, TRUTH, ERROR, AND CRIMINAL LAW, supra note 5, at ch 2.
For English law, see R. EMSON, EVIDENCE 363–366 (1999); P. MURPHY, MURPHY ON EVIDENCE 102 (1980); R. MUNDAY, EVIDENCE 72–74 (2001). Murphy defines very clearly what he understands by standard of proof: “it is the measurement of the degree of certainty or probability which the evidence must generate in the mind of the tribunal of fact” (at 101). Understood in this way, for example, the “beyond reasonable doubt” standard of proof seems to be intimately connected to the principle in dubio pro reo that is employed in civil-law countries. There are obvious differences between these two decision-making rules, but they also have one important thing in common: in each of their respective doctrines and jurisprudence, they are generally interpreted psychologically. The doubt they speak of is the mental state of doubt that, as an empirical question, the fact finder has. And if this is the case, if this is what is relevant, once the fact finder informs us that he or she “has a doubt,” then there is nothing more to discuss. But this leads obviously to an irrational conception of the evidence. A defence of the psychological interpretation of in dubio pro reo can be found in E. Bacigalupo, La Impugnación de los Hechos Probados en el Recurso de Casación Penal. Reflexiones sobre un decenio de aplicación del art. 24.2 CE, 1 ESTUDIOS DE JURISPRUDENCIA (1992), reprinted as E. BACIGALUPO, LA IMPUGNACIÓN DE LOS HECHOS PROBADOS EN LA CASACIÓN PENAL Y OTROS ENSAYOS 68–70 (1994). There is also an interesting comparative discussion between the different standards of proof in common-law and civil-law countries in K. Clermont & E. Sherwin, A Comparative View of Standards of Proof, 50 AM. J. COMP. L. (2002); and M. Tartuffo, Rethinking the Standards of Proof, 51 AM. I. COMP. L. (2003).
Despite its enjoying such a broad acceptance, I consider that there are good reasons to reject this kind of reconstruction based on beliefs. In the first place, it hardly needs pointing out that the truth of a proposition that asserts the possession of a belief on the part of an individual need not necessarily coincide with the truth of the proposition believed by this individual. It is enough to observe that it is possible for an individual to believe that p is true and when p is false. For this reason, if proof of a proposition of fact is linked to the trier having the state of mind called “belief” in relation to this proposition, there arises an absolutely subjective notion of proof far removed from what could be called a rationalist concept of proof (Twining 1994, ch. 3).
In the second place, there are many cases in which, in fact, judges pronounce sentences contrary to their beliefs, not only as a result of the application of legal premises but also in determining the factual premises of their reasonings (Larenz 1960, 302–303). It is not difficult to imagine in this sense that a judge or jury may have the conviction that A has carried out the deed that led to his being accused of crime Y, but that in the light of the evidence brought to trial, the principle of the presumption of innocence must be applied. Likewise, in civil cases, a judge may believe A is the father of B but understand that insufficient evidence has been produced at the hearing. In these cases, clearly, what the judge considers as proven is not based on his own belief. In fact, this situation may arise for at least six different reasons:
- the trier's belief or conviction is of an irrational nature, contrary to the available evidence;
- the trier is in possession of evidence (upon which his belief is based) that has not been incorporated into the proceedings and which therefore cannot be used in his decision (c.f. the classic work of Stein 1893, 133–144, Damaška 1986, 30, 138, 170–171);
- evidence has been submitted (upon which the judge's belief is based) which has subsequently been rejected for legal reasons or for having been obtained in violation of basic rights;
- the decision is based on some legal presumption that has not been successfully challenged by the evidence included in the proceedings;
- the facts are accepted by all the parties as true and therefore accepted as proven even though the judge may not believe they ever occurred; and finally
- the decision entered into the findings of fact adheres to the application of a legal rule concerning proof that predetermines the result of the decision in relation to the facts.
Furthermore, as may have been observed in some of the examples given, the judge not only may but must dispense with his beliefs in order to select the proven facts that he will incorporate in his arguments. This duty is imposed on him by the legal system of reference and is the first element that allows a distinction to be made between the fact that a proposition is proven and the fact that this is accepted as proven by the judge of the case. In effect, in the case of a judge having to abandon his beliefs about the facts of the case in order to determine the proven facts, the judge does not abandon them; it could be said that the judge held that proposition p was proven but that it was not in actual fact proven.
However, we need to analyze in greater detail some of the defining features of belief in order to verify its suitability in the field of judicial proof. In order to do this, I will follow the discussion of Engel (1998 and 2000) and point out two features that seem to make the notion of belief especially unsuitable for explaining the state of mind involved in the findings of fact.
1.
In the first place, we can say that beliefs are something that happen to us, that is, their occurrence in our minds is involuntary (Engel 1998, 143, and 2000, 3 and 9. In a similar vein, see, e.g., Hume, 1739–1740, Appendix to Book III; Williams 1973, 148; Cohen 1989, 369; 1991, 467; Bratman 1992, 3; Redondo 1999, 124–126). Williams's argument in this respect seems particularly convincing: if one could decide to have a belief, then one could decide to have a false belief. However, one cannot decide to have a false belief, therefore, one cannot decide to have a belief. This is owing to the fundamental basis of truth that beliefs have by definition (Redondo 1999, 125). This does not deny that beliefs about the occurrence of an event may be based on evidence, but this basis is simply causal in the sense that if an individual is shown evidence in support of the truth of a proposition, this evidence can cause the belief in that individual that the event occurred. But in no case can the individual in question decide to believe in it (Williams 1973, 141–151). Of course, this does not mean that one cannot decide to act with the goal of acquiring a belief that one does not have. Thus, for example, if I do not know the parentage of King Charles V (Holy Roman Emperor, also known as Carlos I of Spain), I can decide to act in order to find out this relationship (which implies having a belief in it); I can, for example ask a friend who is well versed in Spanish history or I can consult an encyclopedia. Once I have the information, I have the belief that Charles V was the son of Philip I the Handsome and Joanna the Mad. But note that it is not the case that I have decided to have the belief that Charles V was the son of Philip I the Handsome and Joanna the Mad. What I have decided is to take action in order to obtain some belief about the lineage of Charles V.8
For this reason, it makes no sense to impose legally the possession of a particular belief or offer a prize (e.g., money) in exchange for believing in a particular proposition. On the other hand, it has been argued that it would be possible to adopt a belief voluntarily by means of techniques of self-hypnosis; M. B. Naylor, Voluntary Belief, 45 PHIL. & PHENOMENOLOGICAL RES. 434 (1985). But even if this were possible, it is, in my opinion, a clearly marginal case, especially in relation to the discussion of judicial proof. Nevertheless, other authors have also questioned the involuntary nature of belief, among others, B. C. Van Fraassen, Belief and the Will, 81 J. PHIL. (1984) and M. Losonsky, On Wanting to Believe, in BELIEVING AND ACCEPTING (P. Engel ed., 2000).
That said, if this is so, it means that having a particular belief about an event cannot be justified in itself (the content of the belief is something apart), since only voluntary acts admit justification. In other words, if we accept that the result produced by the evidence present in the judicial hearing is the production of a particular propositional attitude about the proposition being proved, it follows that, based on the reconstruction being analyzed, the result may be formulated as “I believe that p.” And if the fact that having a belief is not justifiable in itself, then it becomes impossible to justify this result or, which amounts to the same thing, to justify the trier's evaluation of the evidence. In other words, for those who maintain that the objective of presenting evidence is to produce a particular state of mind (conviction, belief, etc.) in the judge concerning a proposition that they wish to prove (in connection with this, see Ferrer 2002, 72–77),9
D. González Lagier describes this idea as a “fallacy of intimate conviction” as it is based on an unsustainable theory of knowledge; González Lagier, Hechos y argumentos (Racionalidad Epistemológica y Prueba de los Hechos en el Proceso Penal) (I), 46 JUECES PARA LA DEMOCRACIA 39 (2003).
It would be a different matter to expect the justification of the propositional content of the belief. But we would then be justifying not p but rather “I believe in p” and in order to justify p it is irrelevant that p is believed or not by someone, i.e., the judge.
In this context, the most that can be done is to carry out an empirical psychological investigation into the causes that have led the judge to believing that p, that is, to considering that p is proven (Williams 1973, 141–151). And the judge, for his or her part, will be able to justify the verdict, in those systems where it is obligatory, only by explaining (not justifying) the causes that have led him to believe that p.11
In this sense, according to J. MONTERO AROCA, LA PRUEBA EN EL PROCESO CIVIL 28 (1996), “sometimes, the legislators provide regulated values that the trier has to confer to a particular means of proof…. In other cases, the law provides that the trier has to confer a value to the means of proof that he or she deems correct according to the rules of reasoned judgement, in which case certainty is placed into a relationship with the psychological conviction of the judge himself. In this case, we need to talk about ‘subjective certainty,’…since the need to justify the judgement must lead to a demonstration in a reasoned manner as to how a conviction has been arrived at, beginning with the means of proof employed.” Also for this reason, in Anglo-Saxon law, J. D. Jackson, Making Juries Accountable, 50 AM. J. COMP. L. 494–503 (2002), concerned about the justification for the decisions of juries and their accountability, proposed opening the “black box” of the jury's deliberations in order to control whether the reasoning behind its verdict has taken place without prejudice. The problem is with always beginning with beliefs in order to analyze the justification of the decision about the facts.
This consequence seems to be in line with the thesis of those who claim that the free evaluation of evidence by the judge cannot be anything more than the expression of his own inner conviction about the events and is therefore uncontrollable12
This way of understanding the proof has manifested itself in various ways throughout history and today is widely accepted in procedural scholarship and jurisprudence. The idea of “intimate conviction” in its current version comes from post-Revolutionary France (see, e.g., Law 16, September 29, 1791, on penal procedure). On the other hand, it is important to point out that the linguistic formulas already reveal the different ways of understanding the mode of making decisions about facts. In this sense, although one can observe a marked subjectivist hue in intime conviction, this is not the case in the German formula of Freie Beweiswürdigung, which accentuates free evaluation of the evidence (in contrast to legal proof systems). Finally, the Italian formula of prudente aprezzamento appears to underline the quality of reasoned judgment as its guideline for judges in their evaluation of evidence. The system of free conviction, as presented by E. J. COUTURE, LOS FUNDAMENTOS DEL DERECHO PROCESAL CIVIL 146 (1942), becomes especially meaningful for the more subjective definition; Couture asserts that with this formula, the legislator is saying to the judge: “You give your verdict as your conscience dictates, with the trial evidence, without the trial evidence or even in spite of the trial evidence.” An analysis of the various denominations of the system of free evaluation of the evidence can be found in V. DE SANTO, 2 EL PROCESO CIVIL 613–615 (1988). Finally, it should be pointed out that although the idea of proof as a psychological, inner conviction is frequently supported in the theory of free evaluation of the evidence, it is not implied logically from it. Wróblewski, for example, maintains that “the theory of free evaluation of evidence can be summarized as follows: (a) the court should determine the ‘objective (material) truth’ and (b), should use for this purpose all means for obtaining the relevant set of evidential statements; (c) evidence is evaluated free from any legal noms, according to the rules accepted by empirical science and in common experience”; J. WRÓBLEWSKI, The Problem of the So-Called Judicial Truth, in MEANING AND TRUTH IN JUDICIAL DECISION 178 (1979); WRÓBLEWSKI, THE JUDICIAL APPLICATION OF LAW 173–178 (1992); see also M. GASCÓN, LOS HECHOS EN EL DERECHO. BASES ARGUMENTALES DE LA PRUEBA, at 157–161 (1999).
Given that, as Engel declares, we are not responsible for our beliefs; P. Engel, Introduction: The Varieties of Belief and Acceptance, in BELIEVING AND ACCEPTING 11 (P. Engel ed., 2000).
2.
The second defining feature of belief that makes it especially unsuitable for explaining the propositional attitude involved in the findings of fact is its independence with respect to the context (Bratman 1992, 3; Engel 1998, 143–144; 2000, 3; Clarke 2000, 36–38). That is to say, our beliefs are the result of a multitude of factors and information that can progressively change as time goes by. So, in any given moment t, we may or may not believe that p but we cannot believe that p in relation to a context c1 and believe that not-p in relation to a context c2. We cannot, for example, believe or not believe that Hong Kong is a noisy city depending on whether today is Monday or Tuesday or whether we are asked by the mayor or by an environmentalist.14
Although it is perfectly possible to respond differently to one or the other, in one of the cases we would be lying with respect to the actual belief we hold.
I should point out that when the propositional content of the belief includes deixis (such as this, you, now, here, etc.) it becomes obvious that the truth value of this proposition is dependant on the context: thus, for example, the truth value of the utterance “It's raining in Barcelona today” depends on the day the expressed proposition refers to in each case; see, among the abundant literature on the subject, G. FREGE, Der Gedanke, in BEITRÄGE ZUR PHILOSOPHIE DES DEUTSCHEN IDEALISMUS (1918–1919); and J. Perry, Frege on Demonstratives, 86 PHIL. REV. (1977); Perry, The Problem of the Essential Indexical, 13 NOÛS (1979). In these cases, then, the truth value of the propositional content of the belief will depend on the context, but not the actual belief, i.e., not the propositional attitude; M. Bratman, Practical Reasoning and Acceptance in a Context, 101 MIND 3, n. 4 (1992).
The independence from context characteristic of beliefs brings with it serious problems in attempting to give an account of the propositional attitude involved in the findings of fact when there is a disassociation between these findings and the beliefs of the judge concerning the facts of the case. This happens, for example, when the judge or jury has obtained knowledge of evidence that has subsequently been found inadmissible on legal grounds or because it was obtained in violation of basic rights, and so on. In these cases, it is inevitable that the information brought to the process by the rejected evidence will have the effect of shaping the beliefs of the judge or jury concerning the facts. In contrast, such evidence may not be taken into consideration in determining the proven facts of the case. A disassociation of this type can also arise if the trier has out-of-court knowledge of the case that has not been incorporated into the proceedings by any party; it is obvious that such out-of-court knowledge will affect the shaping of their beliefs but in this case also they cannot be used to determine the proven facts of the case.
If this is so, we cannot say in any of the cases described that belief is the propositional attitude of the judge or jury with respect to the propositions they declare proven. If we want to offer a unifying explanation of this propositional attitude, we have to arrive at an attitude that is not independent of context and allows for variation depending on the practical contexts in which the decisions have to be taken.
“It Is Proven That p” as Related to a Fact Finder's Knowledge of p
The previous thesis (“It is proven that p” as related to a fact finder's belief in p), which reconstructs the probative result of the evidence brought to bear in terms of the judge's beliefs about what is considered as proven, is thought to be insufficient by some authors. In order to make this thesis more solid, they include the requirement that the judge knows the facts that he declares proven. In this sense, for example, Bulygin declares:
It is important to realize that facts are what they are and not what judges and other officials say they are. The fact that the fact-finding procedure finishes by an authoritative decision of the judge based on evidence produced in a limited time-span does not alter the fact that this procedure purports to determine the truth and that the judge is under an obligation to search for the truth. (Bulygin 1995, 22)
But one thing is truth and another, quite different, is the knowledge of truth. We say that a person a knows that p, when at least the following three conditions are given: (i) p is true, (ii) a believes that p, and (iii) there is sufficient evidence for regarding that p has been proved. This third point is of utmost importance; for punishing Tom the law requires not only that the sentence “Tom Killed Peter” be true, but also that there be sufficient evidence for the truth of this sentence. (Bulygin 1995, 20)
It should be pointed out here that from the point of view of the judge, that is to say, the person who declares “p is proven,” there is no difference between the requirement of (justified) belief in p and the requirement of knowledge. In other words, a person who believes that p and that the content of his belief is justified necessarily has to believe that he knows that p (Villoro 1982, 129–131). The distinction is, however, important from the point of view of third parties controlling judicial decision. In effect, from the point of view of a third party, it is obvious that he can say that an individual s believes that p, but that p is false and therefore that s does not know that p.
What we now have is a conception that does not lead to subjectivism in the judicial fact-finding. In effect, given that one of the requirements needed to be able to say that a proposition is known is that it is true and that the truth of a proposition does not depend on the will or the beliefs of any individual, what we obtain is a criterion for checking the justification of the judicial decision regarding facts that is independent of the trier: the truth of the proposition declared proven.
Having said that, in my opinion, the truth of a proposition is too strong a requirement in order to consider a proposition to be proven. In other words, I believe that it is plausible that one could maintain an idea of proof according to which a false proposition could be proven. If this point is accepted, then it cannot be maintained that in such cases the judge knows what he or she declares to be proven.
Furthermore, given that the judge's belief in p is one of the conditions required to be able to say that the judge knows that p, we can apply to this second reconstruction many of the objections raised earlier against the requirement of the judge's belief in the propositions he declares proven. That is to say, in all those cases in which the judge can or must decide on the facts of the case contrary to what he or she believes, we still cannot say that the judge is declaring as proven something that he or she knows or believes he or she knows (see Mazzarese 1996, 97).
For all of these reasons, in my opinion, both belief and knowledge ought to be discarded as ways to reconstruct in a general way the propositional attitude implied in formulating “p is proven.”
“It Is Proven That p” as Related to a Fact Finder's Acceptance of p
The first step in analyzing this position is to define what we mean by acceptance of any proposition p. Such a definition is troublesome, and we can hardly say that the word “acceptance” has been used in a uniform manner in the field of philosophy. Nevertheless, with regard to this work, I will use the notion outlined by L. Jonathan Cohen:
to accept that p is to have or adopt a policy of deeming, positing, or postulating that p; that is, of going along with that proposition…as a premise in some or all contexts…. Accepting is thus a mental act. (see Cohen 1989, 368; 1992, 4).16
In philosophical literature there are other designations (with slight variations) used to refer to the same propositional attitude. For example, it is worth mentioning “pragmatic belief” (I. KANT, KRITIK DER REINEN VERNUNFT [1781–1787], reprinted in CRÍTICA DE LA RAZÓN PURA A-824/B-852 [2000]), “assent” (R. De Sousa, How to Give a Piece of Your Mind: Or, the Logic of Belief and Assent, 35 PHIL. REV. [1971]), and “holding as true” (E. Ullmann-Margalit & A. Margalit, Holding True and Holding as True, 92 SYNTHESE [1992]). Given that analyzing the different shades of meaning is not the main aim of this work, I will refrain from further discussion and assume Cohen's notion of acceptance.
There are certain defining features of acceptance I need to point out that are of particular importance in the context of our subject matter. To do so I will return to Engel (1998)
a) In the first place, and this is essential, we are dealing with a voluntary act, that is to say, the person accepting makes a decision regarding his or her acceptance; it is not something that happens to him. This is because the reasons for accepting a proposition, for using it in reasoning, do not have to be epistemic (Cohen 1989, 369; van Fraassen 1989, 192). That is to say, it is not necessary (although in many cases they go together) that a proposition is accepted because it is believed to be true: “[F]or a professional purpose [for example] a lawyer might accept that his client is not guilty even though he does not believe it” (Cohen 1989, 369). In the same way, a salesman may accept the premise that “the customer is always right” and incorporate it into his reasoning even though, for obvious reasons, he does not believe it is always the case. Thus the element of free will grants an essentially practical nature to acceptance (Bratman 1992, 9–11; Ullmann-Margalit and Margalit 1992, 177–179). Furthermore, even when, as we shall see below, the belief in p is the reason for accepting p (i.e., when it is accepted for epistemic reasons), this acceptance implies the decision to accept that which is believed, and there is no automatic step from belief to acceptance.17
One of the reasons this step cannot be automatic is the gradual nature of beliefs, whereas acceptance has a qualitative nature, not quantitative—it cannot be graded; Bratman, supra note 15, at 2; P. Engel, Believing, Holding True, and Accepting, 2 PHIL. EXPLORATIONS 144, 147 (1998).
b) The essentially practical nature of acceptance to which we have already referred, as well as the fact that it does not have to be based necessarily on epistemic reasons (or, in other words, it does not necessarily have pretensions to truth; Engel, 1988, 146), brings with it an additional feature of special importance in the context of the problems I am discussing here. In effect, acceptance, in contrast to belief, is dependent on context (Bratman 1992, 4–5, 9–11; Engel 1998, 147–149). In the words of Stalnaker (1984, 80–81):
A person may accept something in one context, while rejecting it or suspending judgment in another. There need be no conflict that must be resolved when the difference is noticed, and he need not change his mind when he moves from one context to the other.18
Broadly speaking, I believe I have presented a characterization of acceptance that corresponds to the notion maintained by Cohen and that serves as a starting point. Nevertheless, some authors (see, e.g., Bratman, supra note 15, at 11) cast doubts on the fact that, for Cohen, acceptance is dependent on context. In effect, Cohen links this propositional attitude to adopting the policy of using the proposition. Although it is not clear what is meant by “adopting the policy of,” it seems to allow for an interpretation that makes it incompatible with contextual relativity. In my opinion, however, a different interpretation is also possible that is compatible with the contextual dependence of acceptance. I believe—and I go into this in more detail below—that the examples Cohen offers in the judicial field evince a good example of this. In a similar vein regarding the interpretation of Cohen's ideas, see Engel, supra note 13, at 9–10.
If acceptance plays a fundamental role in practical reasoning, it is hardly surprising that one may change what one accepts depending on the context of the decision or for what is being decided.19
This assumes also that the ideal of rational integration is not applicable to the set of propositions that we accept, although it is applicable to beliefs; Bratman, supra note 15, at 4, 9; Engel, supra note 17, at 147.
With these features we can now discuss how useful the notion of acceptance is in attempting to give an account of the propositional attitude generally implied in the judge's or jury's verdict. According to Cohen: “the verdict would declare what the jury accepts, not what it believes” (Cohen 1992, 120). Furthermore:
another important factor that has to be taken into account here is the extent to which the legal system regulates proof. Any such regulation tends to attract attention to the possibility of a divergence between belief and acceptance as the mental foundation for a juror's verdict…. For example, where a proof depends at any point on a presumption…a lay trier of fact may be able to accommodate the presumed truth as a reason for accepting the proposed conclusion but be quite incapable of coming to believe that conclusion. (Cohen 1992, 122; Mendonca 2000, 220, also supports this thesis).
In short, the rules of proof in some cases may oblige the trier to accept a premise in his reasonings, but in no cases is it possible to oblige him to believe it.
If the formula “p is proven” is restructured in such a way as to assume that the propositional attitude of the trier consists in accepting that p, it is then possible to continue asking about the reasons, the justification for this acceptance (it seems also to be understood in this way by Taruffo (1992, 266–272) and Ferrajoli (1989, 42–43, 122–126) who use the criteria of “justified acceptability” for the selection of the factual premises in the judge's deliberations). Legal regulations, in any case, may or may not require the formulation in the verdict of the reasons (i.e., the evidence) that justify the judge's acceptance of p. And it will be on the basis of this evidence that the trier's findings of fact can be evaluated. It may be said in the end that the judge considered as proven (accepted) that p but that in reality p was not proven (its acceptance was not justified) given the evidence existing in the judicial proceedings.
The dependence of the context typical of the acceptance is, therefore, particularly appropriate to explaining the verdicts of trials. It is so because it allows us to translate in terms of propositional attitude an important feature of the declaration of verdicts: their relationship with respect to the evidence incorporated into the judicial proceedings.
1. Belief and Probative Statements Once Again
I have already referred above to the fact that reasons for accepting a factual premise can be epistemic in nature: the trier, in this case, would adopt p as a premise in his reasoning because he or she would have sufficient evidence to accept p. In other cases, in contrast, the reasons for accepting may be regulatory: the legal system, by applying certain rules of evidence, will oblige the trier to accept p as a premise in his reasoning.
When the reasons for accepting a proposition are epistemic, the relationship between acceptance and belief comes up again. Therefore it is worth backtracking for a moment and saying something about this relationship, both generally and also specifically with respect to probative statements.20
I call probative statement a statement that has the form “It is proven that p,” uttered by a judge or a jury in the course of a trial as a part of the judicial decision.
There are many philosophical approaches to the relationship between belief and acceptance. In many cases however, the differences are of nuance. Thus, given that it is not the main objective of this work, I shall not go into an exhaustive explanation of these differences here (a good explanation can be found in Engel 2000).
As pointed out by Cohen (1992, 17), the most obvious relationship between belief and acceptance is that the first can be at the least a prima facie reason for the second. That is, on many occasions, perhaps the majority, we accept a proposition and incorporate it into our reasoning because we believe in it. If this is so, we could say that acceptance is a kind of meta-attitude with respect to belief (Engel 2000, 6). Or, in the terms that I prefer, it can be said that if we assume the belief in p functions as a reason for accepting p, acceptance supervenes on belief (Engel 1998, 148). That said, this does not mean, as I have tried to show above that we should confuse the two.
If we approach this problem by applying it to the case of probative statements, then it is worth making some additional observations. Firstly, a probative statement has the form “It is proven that p.” There are clearly two propositions here: “p” and “It is proven that p,” which are linked metalinguistically. Thus (in those cases where the evidence is freely evaluated, i.e., there is no interfering legal ruling that imposes the result of the evidence evaluation), the probative statement “It is proven that p” must be understood (and expressed by the trier) with descriptive force. It describes the existence in the judicial proceedings of sufficient evidence in favor of p and is, therefore, susceptible of truth or falsehood. And if a statement is susceptible of truth or falsehood, then it must be the expression of a belief (in contrast to prescriptive statements, which express desires). If this is so, we can conclude that in the case of free evaluation of evidence, probative statements are linked to belief. Having said that, it is important to point out this does not mean that the trier's belief refers to p, but rather to “It is proven that p.” That is, belief is linked to the fact that there is sufficient evidence in favor of p. In these cases, the fact finder's belief that p is proven functions as a reason (an epistemic reason) for accepting p. Acceptance of p supervenes the belief that p is proven.
Now, given that the adequacy of the evidence in favor of p is directly related to the evidence gathered together in the judicial proceedings, it cannot simply be said that the belief that p is proven is sufficient reason to believe that p. Thus the trier may believe the former while not believing the latter. If we put this in terms of the truth-value of the propositions, it can be seen that p can be true and “It is proven that p” not true, and vice versa.21
The same arrangement is repeated in statements that attribute propositional attitudes, which are formulated thus: “S believes (accepts, knows, etc.) that p.” Here also, as Frege points out, the truth value of the complete proposition that attributes the propositional attitude is independent of the truth value of the proposition mentioned in it; G. Frege, Über Sinn und Bedeutung, 100 ZEITSCHRIFT FÜR PHILOSOPHIE & PHILOSOPHISCHE KRITIK (1892).
I point out in Section II.A that the proof of a proposition is relational (contextual) to the set of evidence used as a reference. On the other hand, beliefs are inevitably all-things-considered beliefs, and this is one of the most important problems linking proofs and beliefs. So does this alter the belief in which “It is proven that p” coincides with belief in p? And therefore that accepting p coincides with belief in p? The answer is obviously negative for two reasons. The first is that it is purely contingent that anyone accepts and/or believes in p. But if we view the belief and the acceptance as justified, both will coincide if the acceptance of p is based on epistemic reasons and there is no difference between the evidence made available during the trial and the totality of evidence upon which subjects base their justified belief (that is, they do not have out-of-court knowledge upon which to base their belief).
In contrast, in the cases in which the evidentiary value is fixed by a statute (in civil law systems), or when it plays a legal presumption, the reason for accepting p is no longer epistemic; instead it is regulated. Thus, in these cases, acceptance of p is completely removed from the belief that p is proven.
PROOF, TRUTH, AND JUSTIFICATION
To close this discussion on the notions of proof, proven facts, and probative statements, it is worth considering the idea of justification of judicial fact-finding.
First we need to distinguish between the idea that a proposition is proven and that a proposition has been taken as proven by an individual (particularly, a judge or a jury). This distinction runs parallel to another, that is, the acceptability of p and the acceptance of p as empirical data. Thus, continuing on from what is said in the paragraphs above, we can also distinguish between the fact that the judge or jury believes there is sufficient evidence in favor of p and that this belief is justified. In the first case, we can say that the judge or jury believes he or it has (epistemic) reasons to accept p. In contrast, in the second case, we can say that he or it has (epistemic) reasons for accepting p or, what amounts to the same thing, that the acceptance of p is justified or, tout court, that p is acceptable.22
To put it another way, provided that the acceptance is due to epistemic reasons, the acceptance of p can be linked to the belief that there is sufficient evidence in favor of p (but, remember, not necessarily of the belief in p) Here both acceptance and belief are contingent empirical data. In contrast, the acceptability of p is linked to justified belief, or rather the credibility, that there is sufficient evidence in favor of p (and this is independent of the real belief that the trier has in each case). Thus acceptability is not linked to the fact that the judge believes that there is sufficient evidence and that this belief is justified but, rather, to the justification of the belief that there is sufficient evidence, independently of whether the judge believes it or not.
This concept allows us to give an account of the fallibility of the judge or jury in choosing the factual premises in his or its reasoning. In effect, it can be said that the trier erred in choosing these premises if she considered that accepting them was justified whereas in fact it was not, or vice versa; in other words, if the evidence available during the proceedings did not justify adopting the decisions about the facts that were adopted (or if they did justify them but they were not adopted) and providing she did not have a legal obligation to reach the verdict she did because of a particular regulation regarding evidence. Note that what is relevant here is the evidence in favor of p and not the truth of p: as far as the evidence is concerned, the judge will also have reached an unsuitable verdict if he finds that p is proven if there is insufficient evidence in favor of it, even when p is actually true, that is, he reaches the correct verdict by accident.
There is a distinction, then, between the fact that p is proven and the fact that the judge or the jury has accepted that p is proven. The two facts may occur conjointly (which would be a requirement in a rationalist concept of the proof), but either of the two could be so without the occurrence of the other; that is, it may happen that the fact finder accepts that p is proven even when there is insufficient evidence in favor of p or, in the event that she does have sufficient evidence in the judicial proceedings to declare that p is proven, she declares otherwise.23
It should be pointed out here that if the analysis developed in this paper is correct, accepting the premise p and including it in the reasoning allows reconstruction of the judge or jury's reasoning of proof to arrive at her or its decision about the facts (to declare p proved). But it is a contingent fact that the trier of fact accepts p, independently of whether p is justified or not. For this reason, the discourse should be taken one step further and the proof of p should be explained in terms of its acceptability (and not simply of its acceptance; see R. J. Allen, The State of Mind Necessary for a Juridical Verdict, 13 CARDOZO L. REV. 490 [1991]). Acceptability is a notion that belongs to justifying discourse, not to explanatory discourse (used in the preceding sections of this paper). P is proven if there is sufficient evidence in its favor, i.e., if it is acceptable as true. At this point a different and very important problem is encountered: Under what conditions could we say that there are sufficient reasons in favor of p? This is not a question that I will answer here and now. I only want to point out that the sufficiency requirements of the evidence place us in the field of probabilistic discourse and more exactly, according to what I believe, in that of inductive probability.
It is important here to distinguish this supposition from another sense in which a verdict's fallibility might be discussed: in effect, it is possible that the verdict is justified, in the sense that the evidence made available during the proceedings and the procedural regulations of the trial justify the premises accepted by the judge, and yet, in contrast, the events actually occurred differently. In such cases, we can say that the facts declared by the judge to be proven were, in effect, so proven—even though the descriptive propositions of the facts turned out to be false.24
In a certain sense, it could be said that in these cases there was no judicial error. Given that these are cases where the judge is legally (and in some cases rationally) bound to pronounce a certain verdict, the most we can say is that the judicial system has failed to find the truth but never that the judge has made an error in his evaluation of the evidence available. For more, see P. FORIERS, Introduction au Droit de la Preuve, in C. PERELMAN & P. FORIERS, LA PREUVE EN DROIT 12–13 (1981).
In every case, however, the acceptance of p means that p is included in the decision's reasoning as if it were true. That is, p is accepted as a factual premise (whether it is or not). The judicial system establishes in its rules a series of juridical consequences in the event that certain hypothetical events occur. The rules governing proof and proving serve to determine in each case whether such events have occurred. Thus, independently of the existing reasons for accepting a factual premise, this is included in the judicial reasoning as if it were true,25
For more, see Ullmann-Margalit and Margalit, supra note 16.
Wróblewski says in this respect that “a peculiarity of the judicial application of the law [is] that the factual basis of judicial decisions is an existential statement treated as a true statement”; Wróblewski, supra note 13, at 152 (emphasis added).
Nevertheless, even supposing that all of the above is correct, there still remains something unsatisfactory in the reconstruction being defended. Thus the rejection of the conceptual link between proof and truth27
See J FERRER, PRUEBA Y VERDAD EN EL DERECHO ch. 2 (2002).
As is well demonstrated by L. J. Cohen, Why Acceptance That pDoes Not Entail That p, in BELIEVING AND ACCEPTING 55–63 (P. Engel ed., 2000), neither the acceptance nor the acceptability of p implies p, which is what makes the case in this text possible.
M. TARUFFO, LA PROVA DEI FATTI GIURIDICI 46 (1992) is referring to this unsatisfactory feeling when he says that “it seems, in effect, intuitive that the rule would end up unjustly applied and would be, therefore, violated, if the foreseeable juridical consequences for this rule are applied to a case in which the supposed event in question had not occurred.” For more on this, see C. Alchourrón & E. Bulygin, Limits of Logic and Legal Reasoning, in 2 PREPROCEEDINGS OF THE III INTERNATIONAL CONFERENCE ON LOGICA, INFORMATICA, DIRITTO (A. A. Martino ed., 1989), reprinted as Los límites de la lógica y el razonamiento jurídico, in C. ALCHOURRÓN & E. BULYGIN, ANÁLISIS LÓGICO Y DERECHO 312 (1991); and P. Ferrua, Il Giudizio Penale: Fatto e Valore Giuridico, in LA PROVA NEL DIBATTIMENTO PENALE 217 (P. Ferrua, F. Grifantini, G. Illuminati & R. Orlandi eds., 1999).
I believe it is possible to give an account of these contradictory feelings by paying a little more attention to the notions of judicial decision and justification. To do so, I will follow some of the ideas contained in a work by Ricardo Caracciolo.
Caracciolo (1988, 41) points out that the expression “judicial decision” is ambiguous. In one sense, it refers to the act of declaring a judicial resolution, specifically of an individual ruling. In another sense, it refers to the result of this act, that is to say, to the individual ruling announced (less precisely, if you will, to the content of the decision). Thus when discussing the justification of judicial decisions, we have to bear in mind in which of these two senses we are using the expression.30
Ferrua, supra note 29, at 217–223, also distinguishes between the two senses of justification of judicial decisions pointed out by Caracciolo.
On the other hand, according to Caracciolo, the notion of justification is relative. Acts or rules are justified in relation to a set of reasons. Thus we need to ask if the conditions are necessary and sufficient to justify the judicial decision as a ruling, and if they are so, to justify the judicial decision as an act. In terms of a decision-as-ruling, it seems clear that the conclusion arrived at will be justified only if it derives from the factual and legal premises adopted in the reasoning.31
Which corresponds to the idea of internal justification outlined by J. Wróblewski, The Legal Decision and Its Justification, 53–54 LOGIQUE & ANALYSE 412 (1971).
On the other hand, the judicial decision-as-act cannot entertain the same notion of justification as the decision-as-ruling for the simple reason that there are no logical relationships between rules and acts or between propositions and acts. Thus it cannot be said that the decision-as-act is justified if it derives from the factual premises and the rules. One notion of justification that we can use in this case is the idea of compliance or noncompliance with the rules regulating this act (Caracciolo 1988, 44). Hence the judicial decision, in the sense of an act adopting a specific resolution, will be justified if carrying out the resolution is permitted or is obligatory according to the rules governing this act.
In this way, what can happen is that the act of adopting a judicial decision is justified and that, in contrast, the content of this decision, that is, the individual ruling, is not. This explains the two contradictory feelings mentioned earlier because it shows that we can consider as unjustified, for example, that someone is condemned for a crime she did not commit and, at the same time, consider the act or decision to condemn her as correct or justified (in the light of the available evidence).
Furthermore, the distinction outlined above sheds light on another problem, namely, the relationship between substantive rules and procedural rules when justifying a decision (Nesson 1985, 1357). If we take into consideration the decision as an individual ruling, the general rules applied are basically substantive rules. That is to say, those that establish the legal consequences of the hypothetical event that is being judged and that become, as premises, part of the reasoning. In contrast, justifying the judicial decision as an act depends on compliance with the rules that govern the act, that is, predominantly procedural rules.
But if we have before us two distinct senses of “justification” and “judicial decision,” it may be that the act of decision is justified and the content of the decision, the individual ruling, is not; and therefore we can wonder about the juridical effects that are produced by this divergence. Obviously, this is a matter of specific regulation for each legal system, and a general answer to the problem cannot be given. Furthermore, in each legal system there may be no unifying legal provision for all the hypothetical cases that produce unjustified decisions-as-acts or unjustified decisions-as-rules.
In the case of verdicts regarding proven facts, we also need to distinguish between the act of decision and the content of the same. The content consists of a factual premise (or proposition) that is introduced into the judicial reasoning. Following the analysis developed here, the decision-as-act will be justified if the factual proposition is acceptable, that is, that there is sufficient evidence in its favor included in the proceedings.32
A doctrine concerning the sufficiency of evidence, i.e., a theory of evaluation of proof, must necessarily explain the probabilistic character of this justification. And that is precisely what allows a distinction to be made between the two meanings of justification of the decision pointed out (as an act and as content or, if preferable, procedural and substantive justification): the fact that a hypothesis is the most probable, or even that it has high probability given the available set of evidence, does not guarantee that it is true.
Modern legal systems normally establish limits on reopening discussion of the content of decisions—in our case, discussion about the truth of the factual premises. Provision is made for appeals in which this content can be called into question, but once the decision has become final (either because all courses of appeal have been exhausted or because the right to appeal was not exercised), what prevails in law is the general interest that proceedings have a more or less defined time limit and the decision adopted is granted definitive legal status. There remain only a few examples where it is considered that the decision-as-ruling is unjustified (because one or more of the factual premises are false) and so the case must be allowed to be reopened—even though the decision-as-act was justified and the verdict had been granted definitive legal status (Ferrua 1999, 217–218). In such cases, extraordinary appeals are provided for, such as the recurso de revisión (appeal of review) in Spanish law.33
See articles 510 and following of the LEC for the appeal of review in civil law and articles 954 and following of LECr for the same in criminal procedure.
