Why International Criminal Law Can and Should be Conceived With Supra-Positive Law: The Non-Positivistic Nature of International Criminal Legality

International criminal law (ICL) is an achievement, but at the same time a challenge to the traditional conception of the principle of legality (lex praevia, scripta, and stricta – Sect. 1). International criminal tribunals have often based conviction for international crimes on unwritten norms the existence and scope of which they have failed to substantiate. In so doing, they have evaded the objection that they were applying ex post facto criminal laws. This approach, the relaxation of the concept of law by including norms whose existence is doubtful, has apparently served to maintain a concept of strict legality, but it is unsatisfying (Sect. 2). In my opinion, the strict principle of legality that has linked its absolute validity to the positivity of law is not the correct premise. It makes sense to state that positivity and validity do not necessarily go hand in hand (Sect. 3). Applied to ICL, this means that it is neither necessary nor convincing to “conceal” supra-positive law as positive law, as some decisions of the international criminal tribunals do. For this reason, I consider that Radbruch’s formula, consisting in admitting that there are supra-positive limits which positive law must respect in order to be valid, is well-founded (Sect. 4). The path taken by this significant philosopher of law is methodologically convincing, and it squarely faces the problem of the value of positive law. Nevertheless, if we admit Radbruch’s formula and thereby the limited value of positive law (if we claim that the validity of the law depends on it respecting supra-positive minimums of justice), we must also face the problem of the definition of supra-positive values, the epistemological difficulties of having access to them (Sect. 5), and the question of the scope and enforceability of supra-positive law (Sect. 6). In summary, this article aims to explain why Radbruch’s formula offers a convincing conceptual basis for international criminal legality and, in doing so, aims to contribute to the discussion about the foundations of ICL.


International Criminal Law as Achievement and as Challenge for the Concept of International Criminal Legality
In the past, the international community attempted to prevent the atrocities committed during armed conflicts through the well-known international conventions of The Hague and Geneva, which contain the laws and customs of war and the so-called humanitarian international law. Those conventions were addressed to states, obliging them to organize their warfare according to those international undertakings. In this context, citizens of the respective states were regarded by those international law provisions only indirectly, since they were solely submitted to the sovereignty of their state, not to any authority beyond its national borders. However, insofar as these conventions attempted to reduce the violence of armed conflicts, they did not succeed, and -according to some scholars -this lack of effectiveness fed the idea of creating individual criminal responsibility at an international level. 1 Hence, this new kind of responsibility was not the result of reflection on the (possible) relationship between individuals (citizens under the sovereignty of a state) and an international legal order, but of the aim to find another, more promising instrument to reduce the violence of war. That was the background of the well-known proposal of installing an international tribunal to judge the war criminals of the Franco-German war, formulated by Gustave Moynier, as well as the attempt to prosecute German war criminals according to the Treaty of Versailles concluded in 1919. That aim explains (at least, in part) why the International Military Tribunal of Nuremberg (IMTN) had to be installed: after World War II a response to the atrocities of National Socialism was needed and, since a clear and convincing answer to the crimes was necessary and urgent, the IMTN was installed without any in-depth conceptual discussion of its foundations. This atmosphere of urgency has almost always accompanied the history of international criminal law (ICL) -except in the case of the International Criminal Court (ICC) -and it has likewise affected the corresponding conceptual set-up of this area of law. If we observe the history of ICL from the perspective of its results, establishing an international criminal responsibility of individuals can be acclaimed as a great achievement, since it means a juridification of the responses to crimes, and this lends greater legitimacy to such responses. Paradigmatically, the international criminal responsibility established by the IMTN Statute was a more legitimate answer than the summary execution of the German war criminals proposed in the discussions among the Allied powers. 2 However, if we consider the history of ICL from the perspective of the construction of its theoretical foundations, an ICL addressed to individuals implies a real metamorphosis of the relationship between the citizen and 1 3 the international legal order. 3 This substantial element cannot be ignored, drowned out by euphoric attitudes: certainly, ICL is an achievement, but it is also a difficult challenge to explain the foundations of this new international individual criminal responsibility. As George P. Fletcher and David Ohlin noted correctly: "The historical transition from the Geneva Conventions to the Rome Statute also signaled an undertheorized shift from state and communal responsibility to the prosecution of individuals for the same actions that were previously the basis for state responsibility." 4 This under-theorization concerns many substantial issues of ICL and scholarly work in order to provide a theoretical basis to this new area of law that, sit venia verbo, was born and developed in a (justified) hurry. In this regard, the specificity of ICL cannot be reduced to the fact that its sources are not domestic but international (art. 38 Statute of the International Court of Justice (ICJ-St)). Moreover, these new (from the perspective of traditional criminal law) sources of criminal responsibility imply the emergence of a new relationship between the citizen and the international legal order, which has to be explained without the support of a common (stand-alone, universal) international criminal legal tradition. 5 Besides, in this search for the foundations of ICL we should never forget the main driver of its existence: to give a serious answer (punishment) to crimes of the utmost gravity (specially) in cases where the sovereign state that was expected to react instead did nothing, or even protected the criminals and/or was involved in the crimes. This constant aim in the history of ICL is already a reason to formulate the hypothesis that ICL is based on the premise of the primacy of a (minimum) natural law beyond the sovereignty of states, positive bills, and statutes.
One of the substantial issues regarding ICL foundations concerns the principle of legality as an essential principle in traditional CL. In ICL, we may speak of a real relativization of the principle of legality (some scholars even speak of "erosion" 6 ). 7 A significant factor leading to this relativization is that the international criminal responsibility of individuals is based in part on non-written provisions (therefore, lex non-scripta), which do not have the democratic foundations intrinsic to traditional 3 See about the transformation of the traditional idea of sovereignty in the frame of so-called global law, Nieto Martín (2019, 22 ff.). 4 Fletcher/Ohlin (2005, 541, 554). These kinds of "shifts" are usual in the emergence of ICL provisions. Robinson (2008, 946) has noted how international provisions on human rights and other bodies of law were transplanted acritically into ICL. This is unsatisfying, since the nature and purposes of the different bodies of law are diverse (947 ff.). 5 With reason, one of the reviewers of the first version of this article considered it necessary to explain more clearly why I assert the absence of an international criminal tradition. Certainly, the IMNT Judgment was the first step in the creation of a new legal tradition, and the jurisprudence of the ad-hoc tribunals (ICTY and ICTR) as well as the current ICC is contributing to the emergence of such a tradition. However, despite the very important efforts of these tribunals and the doctrine, we are, in my opinion, very much at the beginning of this new tradition and the consolidation of concepts achieved cannot be compared with that achieved in more mature (long!) legal traditions. This is not a criticism but an observation that should be an incentive to continue working intensely on the conceptual apparatus of this (today still) "new" law, the ICL. 6 Bacigalupo Zapater (2012, 58). 7 See Robinson (2008, 927) on the deep tensions in ICL between a traditional legality linked to liberalism and present in the "official narrative," on the one hand, and what Robinson considers a "harsher and more punitive" criminal law conception, on the other hand.

3
criminal law, 8 have vague contours (therefore, lex non-certa) 9 and whose existence may be even unsure. The natural objection against this first statement would be that ICL should not pretend to assume a principle of criminal legality in terms of continental law but admit that the creation of law is largely in the hands of the judges and tribunals, in the way we know from the common law. According to that, in the frame of a common-law-based conception of legality, the vagueness of the sources of art. 38 ICJ-St should be seen not as a problem but as the space in which judges contribute to the creation of law with their decisions. Therefore, the openness of the English law tradition to precedent and most notably to equity (supra-positive law) would suit ICL more than the strict continental legality would do so. Yet, the conditions underlying a common-law-coined ICL do not mirror those of traditional (criminal) common law: common law has a legal tradition which does not exist in ICL. 10 In my opinion, it is precisely this lack of an international criminal legal tradition, along with the reluctance of sovereign states to assume regulations to which they have not agreed, that can explain why ICL has chosen (mainly) the path of conventional ICL: paradigmatically, the Statute for the International Criminal Court (ICC-St). Yet, custom and general principles of law have played a role in criminal proceedings before the international criminal tribunals and even in national criminal proceedings with an international dimension (clearly in the Berlin Wall shootings case), 11 and they therefore deserve attention.
If we put aside the question of the lack of an international criminal legal tradition, which could have helped to compensate for the vagueness of customs and general principles of law (i.e. to guide the judicial creation of ICL), and focus on the way international criminal tribunals dealt with this vagueness, we can ascertain that in the cases in which those tribunals were creating law, since there was no positive law criminalizing the conduct of the accused at the time it was committed, they "presented" their decisions as "acknowledgement" of positive law. As I explore in Sect. 2, there was an interest in explaining the decisions as "mere enforcement" of pre-existing positive law and this led to a "flexibilization" of the concept of "law" in order to incorporate customs or principles of law whose existence was actually doubtful. Otherwise, it would have been necessary to admit that ex post facto criminal law was being applied. 8 Bacigalupo Zapater (2012, 55 ff., 64). 9 Satzger (2008, 139 ss.). See on the fluid character of international custom, Arajärvi (2010, 167). 10 When I say that ICL lacks a legal tradition, I mean it in the following sense: ICL is a young branch of law, with a relatively short history, which establishes a new kind of responsibility, an international criminal responsibility of the individual. In criminal law we count on established legal traditions that provide us with a conceptual apparatus regarding what is understood by crime, what its elements are, what the requisites of authorship are, etc. But there is no international criminal tradition, and proof of this is that the ICL has had to resort to elements of both civil law and common law. The international criminal courts have begun a new tradition in which a conceptual apparatus has to be outlined and built up, nourished by the activity of international criminal courts and scientific debate. I thank one of the reviewers for pointing out the need to clarify this point. 11 In some of the many proceedings against the Berlin Wall shooters, the German courts used the argument of customary international law to deny the validity of GDR legal provisions that covered shootings at the wall. In this regard, Pastor Muñoz (2018, 463-467).

3 2 How International Criminal Courts Have Dealt With the Principle of Legality: The Example of the ICTY
It is worth analyzing how ICL has dealt with the problem of substantiating the punishment of individuals in cases where there is doubt about the criminalization of their actions in international sources at the time of their commission. This was the problem faced by the IMTN and other international criminal tribunals that were installed ex post facto. In order to avoid the objection that they were applying retroactively unfavorable criminal provisions, those tribunals had to "find" an (international criminal) source of law that criminalized that conduct and was in force at the time of the commission of the crimes. In the framework of continental criminal law subject to the principle of lex scripta, the question of whether there is a law criminalizing certain conduct can be resolved with some ease. In contrast, in the framework of ICL, which has unwritten sources (custom, general principles of law), it makes sense to discuss the content and validity of these unwritten norms. Well, it is within this space of discussion that the international courts have developed their argumentation and, I advance here, have "found" the solution to maintaining a strict prohibition on retroactivity in ICL (although, as I will explain below, in my opinion, this solution is not convincing).
Indeed, the first thing that stands out when analyzing international criminal legality is that there are unwritten sources (customs, general principles of law) and that in some cases serious doubts exist about their scope and even their very existence. Such doubts are alarming if we consider that these sources have been the basis of international criminal decisions (for instance, the IMTN Judgment). The problem can be observed firstly in customary ICL, the very existence of which is sometimes uncertain, due to the dubious fulfillment of the elements of this source of law. 12 Customary law is "evidence of a general practice accepted as law" (art. 38 ICJ-St), which means that its emergence requires both a repetitio facti (settled practice) and an opinio iuris (belief that this practice is rendered obligatory). 13 In other words, custom is a generalized practice of the states supported by the conviction of its legal bindingness, an institutionalization of an extended practice. 14 Hence, first is the practice and then the conviction on its bindingness. However, in ICL, the existence of customs establishing the punishability of certain international crimes has often been asserted despite the lack of a practice by states of punishing those crimes. 15 This lack of state practice cannot be remedied by saying the states expressed themselves to be in favor of the punishment. 16 In view of this problem, it is natural to abandon the path of custom and 12 A correct view of this problem in Arajärvi (2010, 163 ff.) 13 See Arajärvi (2010, 166), with references to the ICJ jurisprudence. The opinio iuris is clearly linked to the attitude of states towards breaches: through their reaction towards them it should be shown (I would say: communicated) that the states acknowledge the breach as such and therewith the breached norm. 14 Eichhofer (2007, 7). 15 About the contrary states' practice, Richter (2007, 109 ff.); Dencker (2008, 301); and Renzikowski (2010, 427). 16 Akehurst (1974Akehurst ( -1975. Nevertheless, it is important to recall that flexibilization of the identification of customs is also a reality in the ICJ jurisprudence: Milisavljević/Čučković (2014, 45 ff.) refers to ICJ 1 3 appeal to the general principles of law, which do not require any practice by states, but only the conviction on bindingness. 17 However, apart from the epistemological problem concerning access to those principles, 18 scholars question the legitimacy of founding criminal liability on those principles, 19 especially because of its general nature. 20 Yet, doubts about the existence and scope of non-written ICL provisions have not been openly addressed in international criminal jurisprudence, many decisions instead choosing to tiptoe around the problem. An example is to be found in the decisions of the IMTN 21 and later the decisions of the International Criminal Tribunal for the Former Yugoslavia (ICTY). 22 Of course, it would have been a strong statement to say openly that the crimes to be punished were not criminalized in (written or non-written) positive law provisions, since that would have meant the relativization of the positive law in its capacity to protect the citizen against ex post facto criminal law application. The achievement of the rule of law, namely limiting the state's ius puniendi (its right to punish citizens) through the positivized law, would have ceased being a formal absolute guarantee and would have become a material relativized guarantee, because it could fail in case of priority of other interests. 23 Instead of facing this problem, these courts preferred to construct their solutions on the assumption that there was a prior positive law (lex praevia) providing a basis for the punishment of the international crimes. However, such an assumption was, in my opinion, insufficiently founded. 24 It is convenient to take a closer look at it. jurisprudence in which "enormous discrepancies with regard to the level of thoroughness" in the analysis of state practice can be observed (50). See as well Tomka (2012, 196 ff.). 17 Ambos (2018, § 5, 7); Käli/Künzli (2013, 83). 18 According to Richter (2007, 107), those principles should be deduced from the most "significant" legal orders, where the question arises as to who has legitimacy to decide which legal orders are "significant." 19 Eichhofer (2007, 8). 20 Dencker (2008, 298). 21 See Ambos (1997, 39 ff.) on the prohibition of retroactivity in the IMTN jurisprudence; Renzikowski (2010, 423 ff.), on the application of supra-positive law by the IMTN. 22 The problem of punishment of international crimes without a clear basis in international criminal custom is common to the IMNT and ICTY. It is not a question of doubting the legitimacy of the punishments, but the methodological path taken -to present the punishment as a consequence of positive law (in this case, international custom) when in truth there are doubts about the existence of such positive law (such custom). After the IMTN several relevant international instruments containing provisions on the punishment of international crimes were enacted (as well as the Nurnberg Principles, approved by the UN). However, the existence of international instruments does not mean, in my opinion, that it can be concluded that such international provisions were applicable by the ICTY, because (and I think that in the case of war crimes it is clear) for such provisions to be applicable to the prosecuted cases (example: Erdemovic), it was necessary to prove that at the time of the commission of the facts there were repetitio facti and opinio iuris, i.e., an international custom in force. I thank the reviewer for pointing out the need to clarify why the described problem applies to both IMTN and ICTY. 23 Eichhofer (2007, 8). 24 Weiß (2001, 420 ff.), on the principles of the ad hoc tribunals' jurisprudence.

3
For instance, in many cases prosecuted by the ICTY (especially Erdemović, 25 Mucić, 26 and Ćelebići 27 ) the defendants argued that the ICTY intended to sentence them to penalties higher than those provided for their conduct under the positive law of the former Yugoslavia. The Tribunal replied that, at the time of the commission of the crimes, there were non-written ICL sources providing higher penalties, for the crimes prosecuted, than those provided for by the domestic criminal law of the former Yugoslavia. In support of its argument, the ICTY cited the well-known provision of art. 15.2 of the International Covenant on Civil and Political Rights (ICCPR) ("Nothing in this article shall prejudice the trial and punishment of any person for any act or omission which, at the time when it was committed, was criminal according to the general principles of law recognized by the community of nations"). However, by doing so, the ICTY was not justifying the existence of such non-written ICL. 28 Indeed, the ICTY should have explained which non-written provisions were in force at the time of the commission of the crimes, so that -as the Tribunal wished -the provisions of the ICTY Statute could be seen as an acknowledgment of pre-existing norms. Therefore, when the ICTY stated that "There can be no doubt that the maximum sentence permissible (…) for crimes prosecuted before the Tribunal, and any sentence up to this, does not violate the principle of nulla poena sine lege. There can be no doubt that the accused must have been aware of the fact that the crimes for which they were indicted are the most serious violations of international humanitarian law, punishable by the most severe penalties," 29 the ICTY did not explain nor justify the existence and scope of those non-written provisions. Actually, the same happened in the IMTN, which claimed the existence of non-written provisions (customary law!) criminalizing war crimes, crimes against humanity, and crimes against peace, but did not openly address the following problems: firstly, that the so-called precedent of war crimes actually related to breaches of international humanitarian law that established criminal liability of not individuals but states (international law sensu stricto); and secondly, that there were doubts about even the existence of a precedent for crimes against humanity and against peace in international law. 30 As a matter of fact, in referring to the ICTY decision in the Erdemović case, Kenneth Gallant has stated that at the time of the commission of the crimes there was a custom according to which war crimes could be punished even with the death penalty, because the latter was a usual punishment in the 19th century, it was applied by the Iraqi Special Tribunal for Crimes Against Humanity (IST) and by the Rwanda 28 See the accurate analysis of the Erdemović Judgement by Dana (2009, 857 ff.), as well as Pastor Muñoz (2018, 460-463). 29 Prosecutor v. Mucić, nr. 817. 30 Jescheck (2004, 38-55) considers that the IMTN could refer to customary law only in the context of war crimes (41). Nevertheless, customary law contained provisions regarding state obligations, not obligations of individuals. The IMTN "transformed" international humanitarian law into ICL. By doing so, it established a new kind of responsibility (individual, criminal) and therefore did not apply positivized law, but supra-positive law.
1 3 courts and this has not changed. 31 However, these facts are not enough proof of an actual repetitio facti and the required opinio iuris: Do we really still respond to war crimes with death penalties? Were those penalties in the 19th century of a criminal nature, i.e., were they based on the criminal responsibility of individuals or rather on domestic or military law? Can a single practice, namely of the IST and the Rwandan tribunals, be considered a "generalized" practice? I would say no (therefore, no repetitio facti). In view of this, the question could be raised as to whether the ICTY could have based criminal responsibility on general principles of law of 'civilized' nations. However, it is doubtful that those principles are sufficiently concrete to offer a non-vague definition of crimes, let alone a range of punishments, 32 so that they could hardly serve as a basis for constructing the punishability in the Erdemović case.
In my opinion, both the IMTN and the ICTY should have admitted that, at the time of the commission of the crimes being prosecuted, there were no unwritten provisions founding the criminal responsibility of individuals and, therefore, these tribunals should have expressly said that they were basing the punishment on suprapositive law. As a matter of fact, that is precisely what international criminal jurisprudence does when it invokes general principles of law. As Joachim Renzikowski has correctly pointed out, those principles are an open door for supra-positive law. 33 Hence, by applying those principles tribunals are in fact considering (without saying it) "law" both positivized as non-positivized (supra-positive) law. Therefore, at the end, there is an admission of the limited relevance of the positivization required by the classic conception of legality. This should not be surprising, since there is a clear link between the opinio iuris element (an essential part of the custom and the sole core of the general principles of law) and morality. 34 Besides, one more argument supports the theory of the (veiled) application of supra-positive law by the international criminal tribunals: those (supposed) principles were non-enforceable at the time of the commission of the crimes, since, to that point, there was no court or tribunal with jurisdiction to apply them: a lex praevia without courts entitled to enforce it is (and not by chance) a typical feature of non-positivized law (clearly, of 31 Gallant (2012, 315 ff., 346 ff.), as well as his monograph on the principle of legality in ICL (2009). 32 How then could the ICTY assert that those principles provided higher ranges of punishment than the Yugoslavian criminal law? We could perhaps admit the discussion on the existence of principles that provide the punishability of those crimes (the answer was positive in the Reuter case), but the doubts are overwhelming regarding the possibility of those principles establishing concrete ranges of punishment (the ICTY's answer was surprisingly positive, but unfounded)! See the convincing objections formulated by Dana (2009, 890 f.). 33 Renzikowski (2009, § 85). In fact, it can be said that these principles are the positivization of a minimum of natural law (Finnis, 2020, 1.4). However, it is necessary that the interpreter assumes the existence of a natural law (real, objective) of minimums so that the interpretation of these principles is not left in the hands of ethical subjectivism and these principles constitute an effective limit to the possible contents of the law. 34 Arajärvi (2010, 168) admits a connection between morality and opinio iuris. According to this, the only element of the general principles of law, the sole opinio iuris, would be in direct connection with morality. Therefore, Renzikowski's conclusion would be right.

3
natural law, which is independent of time and history, of states and their sovereignty, and linked to the nature of mankind). 35 In conclusion, in the jurisprudence of international criminal tribunals the usual approach has been to apply supra-positive law without admitting it openly, with such supra-positive law being "presented" as if it were positivized in a custom or a general principle of law. The reality is that, since its inception, the ICL has turned to a supra-positive level to support the punishability of international crimes, even if it has "dressed" such supra-positive law as positive law. And by stating that there is a "law" where the existence of such law is doubtful, the jurisprudence has avoided the reproach of ex post facto criminal law application.
This problem was also raised in the Berlin Wall shootings case, 36 in which, unlike in the ICTY context, German courts had to apply the strict criminal legality principle as defined in the German Constitution (lex praevia, stricta, scripta). In that case, the tribunals attempted firstly to base the punishment on positive law, which proved to be unconvincing, and then to base it on supra-positive law, namely, on the argument of the non-validity of the German Democratic Republic (GDR) legal provision (art. 27 II Border Law) that gave the shootings legal cover. 37 In the frame of that discussion, the German Constitutional Court asserted that the protection provided by the principle of lex praevia applied only if the positive law both had democratic legitimacy and was in accordance with human rights. 38 In doing so, the Constitutional Court acknowledged the relevance of supra-positive criteria and rejected the absolute validity of positive law that has been approved in accordance with the provided procedures. This was remarkable, since it was a statement of a Constitutional Court subjected to a principle of (written and strict) legality, not the statement of an international court like the European Court of Human Rights (ECtHR), which, according to art. 7 European Convention on Human Rights (ECHR), is entitled to base its decisions on non-written provisions. 39 And this statement of the German Constitutional Court is very important for the purpose of this article, since it implies the open admission that positive legality is only a part of legality. The former is limited by supra-positive values, human rights, whether they have been positivized or not. 40 35 Arajärvi (2010, 180 f.) mentions the possibility that the IMTN, by relativizing the principle of legality, could take into account considerations of "the greater good of humanity" or "policy considerations" and concludes that veiled "policy-directed choices" are not compatible with the principle of legality. In my opinion, ethical and political considerations should be distinguished. Only the first would be a correct supra-positive reference point for the relativization of the value of positive law. 36 Vassalli (2010, 78 ff.). 37 See on this case with detail Pastor Muñoz (2018, 463-467). 38 Decision of the German Constitutional Court 24.10.1996. Thereto Alexy (2000, 210 ff.); Vassalli (2010, 110 ff.). Critical Renzikowski (1995, 344, 346); Renzikowski (1992, 154): the prohibition of retroactivity should not be reinterpreted by the tribunals, but when required abrogated by the lawmaker. Also critical: Lüderssen (1991, 482 ff., 485). About the materialization of the formal guarantee of the non-application of unfavorable ex post criminal law: Felip i Saborit (2011, 509 ff.). 39 Streletz, Kessler und Krenz v. Germany and K.H.W. v. Germany, 22-4-2001. 90. 40 Today human rights are enshrined in international conventions, which could be considered a positivization of the law that, in the past, was supra-positive. However, I would like to emphasize here (taking into consideration an interesting comment by one of the reviewers of the first version of this article) that the minimum natural law also operates as a limit when it has not been positivized or when the international 1 3

A Clarification of the Theoretical Premises: On the (Strong Version of the) Principle of Legality and on the Validity of Positive Law
As mentioned, one of the main concerns of ICL is those cases in which the most grave crimes (in a material sense) are permitted (or at least not punished) according to the positive law in force at the time the individual committed the crimes. If, as in the case of National Socialist Germany or the GDR, there is positive law that permits conducts that we consider serious crimes (killing the disabled, shooting unarmed people trying to cross the border), if we assume a principle of the absolute validity of positive law, we must conclude that in such cases the only possible answer is an acquittal. The latter is an unbearable consequence and, in order to avoid it, the right path consists in assuming that extremely unjust positive laws are not law. But this last statement must be based not only on the aim of avoiding the undesired consequences of a strong positivism (in the sense of separation between validity and morality of laws), i.e. the impunity of serious crimes in National Socialist Germany or the GDR, but also on theoretical reasoning. Therefore, before addressing the theoretical solutions to this problem and, above all, Radbruch's proposal, it is appropriate to clarify my opinion on the traditional (strong) conception of the principle of legality, as well as on the relationship between the validity of positive law and its content. In doing so, I do not intend to formulate any new thesis regarding the discussion between positivist and non-positivist positions (or among the different versions of legal positivism), but rather to make clear the conception to which I adhere and the reasons why I do so.
Traditionally, criminal law has been linked to the idea of a strong legality, at least in the framework of continental law: positive law (praevia, stricta and scripta) constitutes a protection for the citizen. According to that conception, only the positive law as a nation's democratic oeuvre (self-legislation) is entitled to define which behaviors (or omissions) are crimes and therefore punishable. The argumentative circle contained in this conception could be considered a kind of closed loop, if its premises were correct. Yet, the premises crumble at so many points that it can hardly be said that the principle is still valid in all its scope. It would be overly ambitious to attempt to discuss this giant topic here. My aim is much more modest, namely to acknowledge two weak points of this conception, which have important consequences for the construction of legality in criminal law and thereby in ICL.
On the one hand, the expectation, associated with the traditional lex praevia principle, that positive law should be capable of completely defining crimes (through Special and General Part provisions) is at least utopic, not only because of the limits of the language as a means of definition of a clear semantic field, but also because the law needs the essential complement of criminal dogmatics (i.e., the criminal law science that establishes the foundations and criteria for imputation of criminal responsibility). The consequences are obvious: firstly, the mere wording of the law treaties that recognize it are not applicable in the territory of a given state. The necessity of appealing to supra-positive law would not lose force if we understood the content of human rights treaties as being constitutive of an international custom of ius cogens and, thus, binding for all states, since the fact that a series of international norms are binding for states does not necessarily mean that they are binding for the citizens subject to the sovereignty of the state concerned. The latter depends on what place international provisions have in the system of sources of law of the state concerned.

3
has neither the ability to define the scope of the crime nor therefore the ability to produce the expected legal certainty (foreseeability). In this sense, the jurisprudence of many constitutional courts has already recognized that no legal certainty exists if its only support is the mere wording of the law. 41 The latter is only capable of producing "semantic certainty." If the law must offer real guidance to its addressees, in the sense of legal certainty, this aim can only be reached by the accumulation of semantic, axiological, and interpretative foreseeability. This means that the protection of the citizen from the state cannot be provided solely by the positive law but by the result of adding positive law, principles, and criminal law science (i.e., dogmatics). Consequently, the law that is applied to the citizen is not fully defined by positive law provisions but by both positive law and jurisprudence, which is the door through which the third element, i.e., criminal law science, enters into the creation of law. In other words, not even the most perfect penal code can offer, through provisions in its Special Part and General Part, definitions that imply foreseeability for the law's addressee as to how that positive law will be interpreted and, thus, what the scope of that law's enforcement will be. The evident second consequence is that the idea of criminal positive law as self-legislation fades because the law to be applied to the citizen is not the sole (democratic legitimated) positive law, but the sum of the latter and the principles and criteria of imputation of responsibility (which do not enjoy formal democratic legitimacy). Thus, positive law is only a (relevant but modest) frame for the construction both of the concept of crime and the norm addressed to the citizen. That is clear in the civil law tradition. However, it would be possible to raise the objection that these reflections are not valid for the common law, bearing in mind that precedent is a source of law and not, as in civil law, mere law enforcement. 42 Indeed, where judicial precedent is a source of law, the interpretative criteria are integrated into positive law, providing the law with greater precision and thus generating greater foreseeability. Now, precedent is not only a source, but also has its material sources, namely, the principles and constructions of criminal science; and such principles and scientific constructions are those which, at the same time, offer a safe framework to define what is foreseeable, in the light of the written laws and, where appropriate, the precedents that may exist.
On the other hand -and this is of special significance for the purpose of this article -it is important to recall the discussion of scholars on the source of legitimacy of positive law and therewith on its validity. Is positive law legitimate because it is positivized, regardless of its content, or does its legitimacy require the conformity of its content with a material idea of justice? For legal positivism, the validity of law (and therefore, of criminal law) does not depend on the morality of its contents. For inclusive legal positivism there would be a place for moral values, but only to the extent that they have been enshrined in the constitution (a material constraint positivized in the Bill of Rights and, thus, a necessary relationship between law and 1 3 morality). 43 Currently, many criminal law scholars question this formal conception of crime. 44 The formal concept of crime is not only unsatisfying but also formalistic in such a way that it turns the lawmaker's task into something arbitrary, into a random decision. This is unsatisfactory because a decision to criminalize is not a mere agreement, a mere convention, like the choice between right-hand versus left-hand traffic! Moreover, from a criminal law perspective, a concept of crime that is subject only to the very broad (and vague) material limits of the Constitution is not convincing. Against a formal (or eventually a "constrained" formal) crime conception, some scholars -whose views I share -conceive the crime in material terms, i.e., in terms of legal moralism. 45 In their opinion, what constitutes a crime cannot be decided by the lawmaker unilaterally and without restrictions, even if he has democratic support and moves within the constitutional frame. Moreover, the lawmaker is not entitled to "define" what is a crime, but only to "acknowledge" which behaviors are regarded by a society as unjust and then to select the gravest of them and define those behaviors as crimes. By doing so, the law becomes a mirror of social identity (of its values) and -how remarkable! -the aim of a self-legislation is thereby satisfied to a higher extent than in the frame of a formalistic conception. The obvious and difficult issue is to determine which social values constitute the identity of a particular society. In this regard, it is important, firstly, to point out that social identity is not necessarily mirrored in the way a society de facto behaves but in its rationality. In this sense, social values are not necessarily reflected in social behavioral habits, but depend on rational convictions about values. Secondly, the subsequent issue is to establish whether that social rationality must be defined from an immanent point of view or if there are external, objective limits to the entitlement of a society to define its values. In this regard, there are three possible paradigms for the definition of unjust behaviors that are relevant precisely in the framework of ICL: (1) the formal definition of crime by the lawmaker without material limits, (2) the material definition of crime by the society in an immanent way, i.e., with the sole limit of that society's rational evaluation of conducts, and (3) the material definition of crime by a society whose rationality is limited by external, objective limits (values). In this article, I will argue that the last paradigm is, in my judgment, the correct one. Therefore, I understand that social values are not defined by a mere agreement, nor by just any rational agreement, but by the expression of a social conviction that respects some minimum material (real) external limits (which constitute immutable values in the sense of the "moral realism"). 46 By admitting the existence of external limits to the possibility of a social self-definition, I am, therefore, not adhering to the postulates of inclusive legal positivism, but to an idea of real (not "agreed") objective values that define the space within which the relevant social definition of values is possible. 45 See Chiao (2018, Ch. 5.2) on the theory of wrongfulness as a requirement for legitimate criminalization (with references to Duff's requirement of a "moral wrong" and Husak's idea that a wrongfulness principle in some form is a categorical requirement for any acceptable theory of criminalization). 46 Therefore, distancing myself from the idea of inclusive legal positivism according to which the rational agreement is a form of objective moralism that allows the renouncing of real moralism, Moreso (2001, 105 f.). See Pastor Muñoz (2019).
Thus, a crime is a crime, independently from the lawmaker's decisions. Therefore, a lawmaker's decision against the social representations of justice, which in order to be valid should be rooted in human dignity (as real value: minimum natural law), does not suppress the criminality of the behavior. The fact that the unjust character of a behavior has its foundation in the mentioned social rationality also makes it possible to justify a different scope for the protection that the principle of legality grants to the citizen. Indeed, when the crime constitutes a malum in se, the flexibilization of the principle of legality is not necessarily alarming, since punishment is based on a certain degree of knowledge of the perpetrator: though their crime was "covered" (= not punished) by the positive law in force at the time of the commission of the crime, they could hardly believe their behavior was right. On the contrary, both in cases of mala in se of less gravity and in cases of mala quia prohibita, the requirements of legality must necessarily be stricter, since in those two cases either the perpetrator was not aware of the seriousness of their unjust conduct (mala in se of lesser gravity) or they did not even have to consider that their conduct could be unjust (mala quia prohibita). Now, there are reasons to consider that a positive law which is materially unjust has, in principle, validity. The reasons have to do with the maintenance of the legal order: 47 if we were to deny validity to any positive law that is, to some extent, unjust (think of an excessive tax law), the authority of the law would be diluted. However, in cases of extreme injustice of the legal system (e.g., a legal system, as occurred in the GDR, allowing the shooting of unarmed citizens who try to cross the border; or a legal system allowing the killing of political dissidents), the positive law lacks validity (in this sense, Gustav Radbruch's formula to which I refer below and, also, John Finnis's view of iusnaturalism 48 ). Therefore, in general, the weighing between the material justice of positive law and the maintenance of the legal order is resolved in favor of the latter: only in the exceptional cases mentioned, i.e. cases of extremely unjust positive law, material justice must prevail. It is precisely these latter cases that have been of particular concern to ICL. Having set forth the premises I assume with respect to the validity of positive law, it is now appropriate to continue with the analysis of legality in ICL.

Why We Should Acknowledge Openly that International Criminal Law is Not Entirely Positivized, or Why Gustav Radbruch Was Right
As mentioned, a historical concern in the activity of international criminal tribunals has been how to proceed in cases in which there is no clear basis in positive law prior to the commission of the crime on which the punishment can be founded. In such cases, international criminal tribunals (IMTN, ICTY) have not, on the one hand, openly questioned the idea of the absolute validity of positive law (i.e. the idea of considering that the basis of punishment or acquittal is constituted by the regulations contained in formal sources of law, regardless of the material justice of the content of such norms), but have, on the other hand, avoided the consequences of such absolute 47 Renzikowski (1995, 337 f.). 48 Finnis (1992, 151 f.).

3
validity. Indeed, if they had been consistent with the premise of the absolute validity of positive law, in many cases they would not have been able to find a positive law basis for convicting the perpetrators of international crimes. 4950 The path they have taken to avoid the consequences of the absolute validity of positive law, without renouncing this premise, has been to affirm, without substantiating, the existence of unwritten international criminal provisions underlying their decisions, whereas the real basis for their decisions has been supra-positive law. By doing so, they intend to avoid reproach for violating the prohibition on the retroactive application of criminal law. 51 In my judgment, this solution, namely "concealing" supra-positive law as positive law, produces the appearance of maintaining both a strict paradigm of the absolute validity of positive law and a strict formal prohibition of retroactivity but, in truth, it relaxes the concept of law itself, making it so vague that it can be adapted to the necessities of punishment. Methodologically, this path is highly unsatisfying. In the scientific discussion dealing with the problem of crimes that are covered by positive law, some scholars defend the absolute validity of positive law; 52 others have proposed the general abrogation of the prohibition of retroactivity in ICL. 53 Yet, the first solution has consequences difficult to assume; the second contradicts the international conventions on human rights (art. 15 ICCPR; art. 7 ECHR), which some scholars consider even to be universally valid, 54 and is thereby a self-contradiction. Besides, it seems arbitrary to abrogate generally, without a material justification, the protection provided by the prohibition of retroactivity.
In contrast to these two paradigms as well as to the argumentation of the international criminal tribunals, Gustav Radbruch's formula showed a fourth path that was not only honest but solidly grounded from the perspective of the philosophy of law: the relativization of the validity of law, specifically, the denial of validity to that positive law which enters into an intolerable conflict with justice. It was not only Radbruch who openly pleaded for a relativization of the validity of positive law, but also other German scholars of his time like Helmut Coing,55 Hans Welzel,56 and 49 This would have been the consequence of assuming an absolute validity linked to the positivistic paradigm: Grünwald (1971, 14), p. 14; Gropp (1996, 396). Of course: Hart (1957Hart ( -1958. A deep analysis of his opinion by Neumann (2012, 287 ff., 293 ff.). 50 Or to base the conviction on moral reasons, if it is admitted that the international criminal tribunals are morally authorized to apply moral rules (so Finnis, 2020, 3.1.1. on the position of exclusive legal positivism regarding the IMNT). However, this path would mean denying the principle of the prohibition of retroactivity of criminal laws. 51 The introduction of the second section is linked to the justification of the Nuremberg Judgment: Gil Gil (2010, 131 ff., 133). 52 Grünwald (1971, 14). 53 References in Gallant (2012, 337). See also Zappallà (2003, 196). 54 Gallant (2012, 315, 320 ff.); Dana (2009, 873 f., 878 f.). 55 Coing (1947, 61 ff.) understood justice as natural law-based and the function of law as the realization of justice. In his opinion, the duty to deny obedience to unjust positive laws is not of a legal but of a moral nature. Arthur Kaufmann (2000, 33) considers that Coing's conception of natural law goes too far, since natural law does not contain concrete provisions but only -in the sense of Thomas Aquinas -abstract principles of justice and morality. 56 Welzel (1949, 2), in whose opinion a legal provision that degrades the person to a mere thing is not a binding law, since it does not respect the minimal requirements of the order of a community. Here Welzel 1 3 Heinrich Mitteis. 57 Gesetzliches Unrecht und übergesetzliches Recht (Statutory Injustice and Suprastatutory Law): Notwithstanding that in case of conflict between legal certainty and justice legal certainty should prevail, if positive law contradicts justice in an unbearable way, justice has to prevail; therefore, this extremely unjust positive law is not law (intolerability formula). 58 Some have said that Radbruch's rejection of legal positivism (see as well the "Five Minutes of Philosophy of Law" 59 ) and the corresponding loss of relativism 60 had its roots in the experience of National Socialism and its atrocities, 61 and this could indeed be a part of the explanation of Radbruch's formula. But the other part of the explanation is that, since his entrance into Neo-Kantianism, Radbruch had been searching for a connection between law and value (until the point of breaking the methodological dualism and defending the so-called methodological trialism). 62 This question is complex, since we do not know for sure if Radbruch was a positivist before the war. 63 At that time, his concept of law was founded in three elements that were at the same level: justice (understood as formal equality), purposiveness (Zweckmäβigkeit), and legal certainty. 64 Depending on the interpretation of his pre-war conception, there are two opinions on the meaning of the irruption of the formula: the turn thesis and the accentuation thesis. According to the first, Radbruch, a Saul, became a Paul, 65 since he abandoned the relativism of values, the Wertrelativismus (Robert Alexy, 66 Renzikowski 67 ). According to the second interpretation, the formula meant only a change of accent in Radbruch's conception: seems to suggest the existence of a minimal natural law. However, in other texts he defends a link between supra-positive law and the spirit of the times (Zeitgeist: changing supra-positive law). 57 Mitteis (1947, 121 However, leaving aside the discussion about Radbruch's evolution, the present purpose is to analyze the scientific solidity of Radbruch's formula, a statement on the value (and validity) of positive law that is applicable to any positive legal system (the examples Radbruch uses refer to domestic German law, not international law, and not even criminal law). The essential elements of the formula can be summarized as follows: firstly, supra-positive law is superior to positivized law and is, therefore, the criterion to decide if the latter is "real law" or only "apparent law" without validity. Therefore, validity and bindingness do not derive from the observance of formal procedures in the enactment of positive law; in order to be valid, positive law cannot deviate wildly from the idea of justice (consequence: not formal, but material legitimation of positive law). Secondly, the definition of justice is not the product of the lawmaker's decision (even if the latter is a democratic lawmaker). Thirdly, a positive law judged as "extremely unjust" does not bind the citizen or the judges: hence, it does not establish a duty to obey. In this regard, Radbruch converges with the thesis of Saint Augustine, according to which a state without justice is a band of robbers. 72 Such a state represents the perversion of the sense of law, which is to guarantee the rights of the citizens. In other words, that state exerts "pure coercion." 73 The path of Radbruch's formula, followed by the German Constitutional Court in the Berlin Wall shooting case, is the methodologically convincing way to open positive law to suprapositive values. And this is actually what lies behind the reference to the general principles of law as a source of law in many international conventions (art. 7.2 ECHR: in Renzikowski's words, the place where natural law breaks into ICL 74 ). Moreover, through the acceptance of those principles as a source of law, ICL has to accept that it is defining itself as non-positivistic. 75 68 Wolf (1973). 69 Arthur Kaufmann (1987, 27 ff.). 70 Neumann (2007 I, 11 ff.); Neumann (2007 II, 381 ff., 385 ff.). See too Saliger (1995, 31). 71 Paulson (1995, 490, 493 f.) sees the formula as the "correction of a mistake" in Radbruch's pre-war papers. Critical with Paulson's interpretation Spaak (2009, 277 ff.) who considers that Radbruch's pre-war thesis of judicial bindingness is not compatible with his post-war rejection of the separation thesis. 72 In this direction, Coing (1947, 61 f.), in whose opinion the judge's duty to apply the law is neither blind nor absolute: he is only bound in absolute terms by positive law not linked to justice questions; however, he may and must deny obedience to unmoral positive law. 73 See Brix (2006, 143 ff.) on the discussion about whether Radbruch's formula is a theory of law and/or contains instructions for judicial decision-making. Brix seems to see Radbruch's formula as a statement about judicial decision-making, but casts doubts on its nature as a theory of law. Here, on the contrary, and following Alexy, Radbruch's formula is considered as a strict theory of law with consequences for judicial decision-making. 74 Renzikowski (2009, § 85). 75 One of the reviewers suggested the interest of differentiating the position defended here with respect to that of Ronald Dworkin, a suggestion that seems to me to be very appropriate. It is true that Ronald Dworkin opens a door to morality in the law by saying that moral standards function as a direct source of law. However, he considers that when these standards fit in a very weak way with the set of social-fact sources of the community, the judge applying them is applying morality, not law. The iusnaturalist approach I'm assuming would consider, on the contrary, that the judge in that case (let's say Nurnberg, for instance) is 1 3 In my judgment, this path is honest and consistent, since it makes the law's validity dependent on its material justice (the respect of a minimum of natural law) and hence relativizes the (legal) duty of obedience in the case of extreme injustice of positive law. Regarding this last aspect, Renzikowski has pointed out that relativizing the duty of obedience leads to the consequence that the maintenance of a "state of law" (rechtlicher Zustand), i.e. the conditions that make it possible for law to fulfill its functions, cannot be guaranteed. According to Renzikowski, the "state of law" would be jeopardized if citizens were entitled to question the validity of legal provisions and decide on their bindingness. Therefore, in his opinion, in the frame of an unjust legal system, the duty of obedience remains intact, and only a retroactive application of a posterior law is allowed. 76 In my judgment, the relativization of the duty of obedience is actually a delicate matter and the dangers denounced by Renzikowski should not be ignored. It is not by chance that the rule of law is seen, by iusnaturalists too, as a remedy for the dangers in having rulers. 77 Notwithstanding that, maintaining an absolute duty of obedience, even in the case of an extremely unjust positive law, leads to the undesired consequence that judges unwilling to apply that unjust positive law commit prevarication. 78 In fact, this is the place to recall the main objection raised against the positivistic conception of the duty to obey positive law 79 : a duty of obediapplying law, not morality, since some minimal moral standards (values) are (for the naturalist point of view) law (see Finnis 2020, 3.2). In this sense, Radbruch's premises (at least in his formula) are in my opinion substantially different from Dworkin's position, because in Radbruch's formula, supra-positive law is law. Precisely because of its nature as law, supra-positive law can "abrogate" extremely unjust positive law provisions. 76 Renzikowski (1995, 337 f.). 77 See Finnis (2020), 1.3. 78 To consider that they were committing a crime does not bar the possibility that some other elements (for instance, coercion in a hierarchical structure) could eventually play a role concerning their grade of responsibility. 79 See on the foundations of the duty to obey in Hart's positivism, Mertens (2002, 194 f.): in Hart's opinion, law's validity is independent of its accordance with society's morality (strong separation thesis). Nevertheless, Hart's conception of the duty to obey has some differences compared to Austin's conception, since Hart separates validity and duty of obedience. According to Hart, citizens disobeying morally offensive rules disobey valid law, but based on moral reasons, not on pragmatical ones (on the classification of the problem of obedience of unjust law as a moral dilemma see Fuller, 1958, 633). Mertens (196 ff.) says correctly that Hart does not provide a good reason why those "moral claims" should not be considered by legal theory. Moreover, there are other (non-direct) points of connection between law and morality in Hart, which has led Moore to claim that Hart's positivism is softened (Moore, 1998, 306): the condition of law depends on two elements, namely, that rules of behavior are generally obeyed and that the "officials" (the judges) have an internal attitude toward the rule of recognition: "Hart locates law's normativity in judges who regard themselves as obligated by the rule of recognition of the legal system in which they are judges" (Moore,307). Indeed, Hart does not take the turn of requiring that the law must be just for judges to develop that internal attitude, but one might think, as Moore has expressed, that normally (though not necessarily) a rule of recognition (the Constitution) will be accepted by judges if its content is seen by them as just; therefore, in Hart's positivism there could (not should) be a limited acceptance of some normativity in the law (Moore,307). Now, note that the most that can be drawn here is that in Hart there may be a connection between law and the subjective morality of judges, but not a connection between validity and objective (and real) justice in the sense of iusnaturalism. As Finnis (2007), 41 f. has concluded: "In short: The Concept of Law, the Essays on Bentham, and Parts I-III and V of Essays in Jurisprudence and Philosophy display a legal theory or general jurisprudence that, having identified its own descriptive dependence on the internal point of view and attitude (in which rules are reasons for action), leaves those reasons largely ence defined in such terms is a "coercion duty." Moreover, it is contradictory to call it a "duty," since, in Welzel's words, it obliges as much as does the pistol of the robber on the temple of a bank employee. 80 Having said that, the obvious tension between justice, on the one hand, and legal security, on the other, requires a deeper analysis of Radbruch's formula in order to see whether the costs of its admission are bearable.

The Concept and Sources of Supra-Positive Law and the Epistemological Problem
Accepting a supra-positive order of values (supra-positive law) as being the source of legitimacy of positive law means facing some difficulties concerning the concept and sources of supra-positive law and the epistemological access to it. As a matter of fact, Radbruch rejected the concept of law being value-neutral and conceived it as a value-bound law that must mirror the right values. 81 Therefore, since he intended to bind the lawmaker (including the democratic one) to a superior order of values, the assumption of Radbruch's paradigm requires an explanation of both the sources of those supra-positive values and the epistemological access to them.
What is the source of supra-positive law? Radbruch did not explain that explicitly, though it seems that he assumed a minimum of natural law 82 or at least some supra-positive legal ethical limits (using Arthur Kaufmann's term: "negative natural law"). 83 The two main options regarding sources of supra-positive law are as follows. The first consists in defining supra-positive law as a universal law which is permanent and valid independently of time and history. Obviously, the assumption of this paradigm requires a real and objective conception of value (if values were the result of a construction of the cognizant subject, there would be no place for values in the sense of moral realism), as well as its permanent nature. In this regard, Jesús-María Silva Sánchez has argued recurrently that external limits to criminal law, derived from natural law, are necessary: an ontology of the person as "dignified essence," unexplored, and rests largely content with reporting the fact that people have an attitude which is the internal aspect of their, practice. Having so fruitfully gone beyond the observer's or spectator's perspective on bodily movements and behavior, it rests officially content with a report that the participants have reasons for their behavior and their practice. It does not seek to understand those reasons as reasons all demand to be understood -in the dimension of soundness or unsoundness, adequacy or inadequacy, truth or error." Moreover, we may say that Hart endorses an ethical subjectivism (Finnis, 2007, 47). 80 See Welzel (1959, 833 ff.): a deep criticism of the positivistic conception of the duty to obey as well as of the theories of the factual validity of law. See also Engisch (1968, 60 ff., 66 f.) and Arthur Kaufmann (1987, 72). 81 Radbruch (1956, 34 ff.): law is the reality which has the purpose to serve the value of law, the idea of law. And the idea of law is, in Radbruch's conception, justice. Nevertheless, Radbruch's conception of justice is a formal one (formal equality), which is logical if we consider the relativistic frame of his Philosophy of Law. Therefore, the formula means a materialization of the idea of justice. Alexy (1992, 8 ff., 17 ff.) points out that, despite Hart's reproach in "Positivism and the Separation of Law and Morals" (1957)(1958), Radbruch does not identify law and morals, but establishes only a minimum of moral limits that positive law should observe to be valid. 82 Engisch (1968, 67 f.).

3
"truth," even if there are substantial difficulties for the epistemological access to that truth. 84 The second interpretation defines supra-positive law as an order of values that changes with the "spirit of the times" (Zeitgeist), 85 namely, that depends on the general (shared) representation of a concrete society in a concrete moment on the non-disposable minimal requirements of human dignity. In this direction Hans Welzel maintained that there are no permanent and generally valid propositions about what is social-ethically right, but only propositions valid for a concrete time and society. Hence, in his opinion values depend on the state of discussion of each generation and cannot be defined in an aprioristic way. 86 Here, values could be considered "objective" in the sense claimed by some positivists, namely, as the "objective" result from a rational agreement (i.e. objective moralism, but no real moralism). 87 Within this paradigm, defending an immanent solution, Bernd Schünemann has said, on the occasion of analyzing Radbruch's formula, that laws are not the "dead letters of a printed legal provision" nor the "arbitrary behavior of a power-clique" but a "public communication of rules" compatible with the culture of a society, and concluded that the unjust positive law has validity if the society accepts its provisions, no matter if we consider (from an external point of view) that law unjust. 88 The problem of this second solution is that it does not escape from relativism, so that, when the "spirit of the times" is incompatible with minimum values linked to human dignity, this solution does not offer a possibility to question the validity of laws that are in an immanent sense (according to the Zeitgeist) just, but unjust from the perspective of the minimums of natural law. Therefore, I am more convinced by the first solution, despite the difficulties it implies because it will not be accepted by those who deny real moralism and because of the epistemological difficulty in accessing these minimums of natural law.
In my opinion, there are some clear premises for the task of determining suprapositive law. Firstly, a paradigm of absolute relativism is not appropriate to achieve a definition of supra-positive law, since a relativism without some objective limits leads to the absolute freedom of the subject in the definition of values (moral subjectivism), which means that their decisions (the same concerns the lawmaker's decisions) cannot be submitted to external control. In other words, without a nonrelativistic point of reference, it is impossible to establish an epistemological relationship. 89 This problem is attenuated, but not eliminated, when values are conceived as rational agreements. These are certainly objective, but they do not guarantee a link with justice, unless we include in "rationality" the limit of immutable external values. Secondly, a conception of a supra-positive order as a concrete legal order is not a path to go down, since it would mean ignoring the historical (changing) dimension of 84 Silva Sánchez (2010, 283 f.): ontology limited by epistemology. 85 Robles Planas (2012, 20); Robles Planas (2010, 357 ff.). 86 Welzel (1959, 835). 87 Moreso (2001, 106 ff.).

3
law. 90 Thirdly, contrary to those scholars who pretend that natural law can be a complete legal order, supra-positive law can only be considered a lex imperfectae 91 (not providing concrete legal consequences, but only general values), so that no concrete punishability (in terms of the scope of the punishable behavior as well as the range of punishments) can be directly derived from that supra-positive law. In my opinion, the right way consists in admitting minimal, immutable (real) values linked to human dignity, leaving their concrete meaning and the legal consequences of their violation in the hands of the respective historical context. The scope of this permanent problem of philosophy of law is too large to be solved in this article and is still the subject of discussion. However, it must be emphasized that, in my opinion, it makes sense to start from a relationship between law and (real) morality which can be relevant (in Radbruch's sense) to assert or deny the validity of positive law.
The most complex issue is the epistemological access to the minimal supra-positive values that positive law must respect in order to be valid. Indeed, in my opinion, the values that bind positive law do not depend on the outcome of formal processes (the majority opinion expresses an agreement, not necessarily the ascertainment of real natural values), or on the way of life of a society (this belongs to the Sein, not to the Sollen: we do not necessarily behave as we should behave), or on a "social rationality" that conceives itself as non-limited, but on a social rationality subject to minimum supra-positive values. However, this conception confronts us with the task of defining those values that operate as constraints on the "social freedom of self-definition." In this regard the argument of "evidence" can provide access to an essential nucleus that to anyone's eyes seems indisputable (strictly speaking, evidence would have to do with a universal consensus and, therefore, with a knowledge that stems from human conscience). However, this path leads only to a very basic nucleus of supra-positive values. As soon as we abandon that core, the argument of evidence is no longer valid and we must resort to other epistemological paths.

Scope and Enforceability of Supra-Positive Law
Notwithstanding the solidity of Radbruch's formula, its consequences for positive law can be of great importance, namely, the denial of validity of any positive law that unbearably diverges from the idea of justice. This implies that there are two essential questions, namely, one concerning the scope of the formula and the other concerning the possibility of its enforcement by courts. On the one hand, it seems of utmost importance to determine its scope of application, since the random use of it could lead to the destruction of the conditions of respect for the law that are necessary for social coexistence. Radbruch conceived his formula as an exception to the general validity of positive law: according to Radbruch, positive law loses its legal 90 Arthur Kaufmann (2000, 24): Thomas Aquinas did not conceive natural law as a complete legal order, but as an abstract order the concrete contents of which are changeable. With more details on Aquinas' opinion on the non-validity of unjust positive law (corruptio legis): Seoane (2006, 316 f., 318). 91 Jakobs (1994, 1 ff.); Pawlik (1994, 472 ff., 481).

3
nature only in cases of unbearable contradiction between law (statute) and justice (therefore, the formula establishes a weak connection between law and morality). But when is the injustice of a positive law "unbearable"? We will probably agree in the core cases (it is unjust to kill children, or sell them for prostitution or use them as soldiers, or to rape), but will come to dissenting conclusions once we leave the core crimes and evaluate less evident offenses. 92 Therefore, the idea of "evidence" will apply to some core cases. 93 On the other hand, the enforcement of supra-positive law by international criminal tribunals seems at first sight less problematic than in the case of domestic legal systems with a strict legality principle, 94 because international tribunals work with a more relaxed conception of criminal legality including custom and general principles of law, which leaves a door open to supra-positive law. 95 Notwithstanding that, the enforcement of Radbruch's formula by the afore-mentioned tribunals poses some questions, because supra-positive law is a lex imperfectae that establishes neither consequences (i.e. the range of punishment) for the violation of its values nor, obviously, provisions on jurisdiction. 96 Hence, it is easier to rely on supra-positive law for the abrogation of unjust positive law provisions (for instance, abrogation of justifications like that provided by GDR-law for the Berlin Wall shooters) than to base on it the punishment of crimes not embraced by positive law. 97 In case of abrogation, the consequence is, firstly, that the non-valid positivized provisions do not guarantee impunity and, secondly, that the enforcement of the unjust provision by the judges is unlawful (although there can be other reasons to renounce or reduce punishmentfor instance, coercion). If punishment is founded on supra-positive law (criminalization), the difficulties are more substantial, since that law provides neither a concrete description of the scope of the crime nor a range of punishment (lex imperfectae and vague! 98 ). Going back to IMTN and ICTY, those tribunals based the unlawfulness of the behaviors on supra-positive law (without admitting it openly), but since the latter did not contain provisions on the legal consequences, the range of punishments applied was indeed based on ex post facto law (the corresponding statutes). There-92 In my judgment the core of values is not to be identified only with actual international crimes, but should be extended to others, since the reason for criminalization in the ICC-St is not only the utmost gravity of the behaviors but also its "international dimension" (not to mention the role that political pressure has played in the process of defining international crimes). 93 It is important to remark that once the clear unlawfulness (according to supra-positive values) of the behavior and thereby the necessity of relativizing the validity of the positive law which granted impunity has been established, an additional question arises that will not be analyzed in the present paper: the question whether it is possible to weigh between the fact that those supra-positive law based crimes deserve punishment and other elements related to state stability, social peace, etc. In other words, the discussion about the admissibility of amnesty laws, i.e., the problem of the possibility of a state's entitlement to renounce punishment in order to maintain social or political peace. 94 See Pastor Muñoz (2018, 455-487). 95 Renzikowski (2009, § 55). 96 Jakobs (1994, 13). 97 Alexy (1993, 107 f.) considers that the formula is only to be applied in order to abrogate unjust provisions, not to criminalize. Renzikowski (1995, 345), in contrast, sees no more than a technical difference between abrogation and criminalization. 98 Koch (1977, 164).

3
fore, concerning punishment, the formula is unable to instruct those judges working in the frame of the unjust law system as to the legal consequences they should apply to the crimes that were not criminalized by the unjust positive law.
Having said that, the reader could wonder if, in the end, Radbruch's formula leads to the same consequences in ICL as those achieved by international tribunals applying "general principles of law," in which case, why place so much importance on Radbruch's formula? Much ado about nothing? Certainly not, since Radbruch provided the relativization of the value of positive law with clear theoretical foundations. That is his outstanding contribution that I have tried to emphasize in these lines. However, despite the enormous theoretical value of Radbruch's contributions, a final doubt may arise: does it make sense to continue to invoke Radbruch's formula in the framework of the ICC, which is bound by a strict principle of legality, as provided for in arts. 22 to 24 of ICC-St? The answer must be positive, for it should be recalled that the ICC can operate both as a tribunal with jurisdiction prior to the commission of the act, and as an ad hoc tribunal ex post facto, i.e., as a tribunal that had no jurisdiction to prosecute the crimes at the time they were committed. In this sense, the ICC as an ad hoc tribunal is driven by the same logic as the IMTN, ICTY, and the International Criminal Tribunal for Rwanda having to deal, therefore, with the same problems of legality that the latter faced.
is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/ licenses/by/4.0/.