Unlocking Legal Validity: Some Remarks on the Artificial Ontology of Law

  • Paolo Sandro
Open Access
Part of the Law and Philosophy Library book series (LAPS, volume 122)


Following Kelsen’s influential theory of law, the concept of validity has been used in the literature to refer to different properties of law (such as existence, membership, bindingness, and more), and so it is inherently ambiguous. More importantly, Kelsen’s equivalence between the existence and the validity of law prevents us from accounting satisfactorily for relevant aspects of our current legal practices, such as the phenomenon of “unlawful law.” This chapter addresses this ambiguity to argue that the most important function of the concept of validity is constituting the complex ontological paradigm of modern law as an institutional-normative practice. In this sense, validity is an artificial ontological status that supervenes on that of the existence of legal norms, thus allowing law to regulate its own creation and creating the logical space for the occurrence of “unlawful law.” This function, I argue in the last part, is crucial to understanding the relationship between the ontological and epistemic dimensions of the objectivity of law. Given the necessary practice-independence of legal norms it is the epistemic accessibility of their creation that enables the law to fulfill its general action-guiding (and thus coordinating) function.


Kelsen Existence of norms Formal validity Material validity Unlawful law Ontological objectivity Epistemic objectivity Normative ontology Practice-independence Binding force 

1 Introduction

Can anything interesting at all still be said on the topic of legal validity (“validity” hereinafter)? Many scholars would reply negatively, given that this concept has been at the forefront of the jurisprudential debate for the last few decades, in particular as one of the elected “battlegrounds” between legal positivists and nonpositivists. The real scope of this debate has been far from clear, however, for what positivists and antipositivists have been debating are the “grounds of validity,” rather than the concept of validity per se (Sartor 2000). Does this suggest then that any form of consensus on the concept has been reached? Hardly so, as shown by the different contributions in this volume.

As we shall see, due mostly to what is a major misconception in Kelsen’s influential theory, “validity” has been used to refer to different properties of law. The result is that the use of the concept in legal discourse is inherently ambiguous, and so it is still unclear what is at stake when we discuss it. My aim in this chapter is to dispel this ambiguity in order to reveal what I see as the most important function played by the concept of validity, namely constituting the complex ontological paradigm of modern law as an institutional-normative practice. Far from being “a mere summary-concept” (von der Pfordten, in this volume), validity is an artificial ontological status that supervenes on that of existence of legal norms, thus allowing law to regulate its own creation. This function, I argue in the last part, is crucial to our understanding of the relationship between the ontological and epistemic dimensions of the objectivity of law. Given the necessary practice-independence of legal norms, it is the epistemic accessibility of their creation that enables the law to fulfill its general action-guiding (and thus coordinating) function.

The chapter proceeds as follows: in Sect. 2, I analyze the source of ambiguity in Kelsen’s equivalence of existence, validity, and bindingness of law and show its counterintuitive consequences vis-à-vis our current legal practices. In Sect. 3, I offer some remarks on normative ontology and in particular on the difference between the epistemic and metaphysical conditions of existence of norms. This allows me, in Sect. 4, to illustrate the difference between the validity and the existence of legal norms and thus to account coherently for the mysterious status of “unlawful law.” Lastly, in Sect. 5, I show the function played by validity within this complex ontology and how it is ultimately connected to the objectivity that characterizes our modern legal practices.

2 The Ambiguity About Validity

2.1 The Source of the Ambiguity

“Validity” appears to be one of those concepts for which it is easy to provide examples—“this contract is valid”—but whose definition proves a considerably harder task. While in general parlance the property of being “valid” indicates the state of being in accordance with a set of criteria (Beltrán and Ratti 2010), thus distinguishing between “good” and “bad” tokens of a certain type or between members and nonmembers of a certain set, the use of the concept in legal discourse has proved difficult to account for consistently. Carlos Santiago Nino (1996, p. 117), to mention but one, has identified at least six different “cores of meaning” in which validity is used1:
  • as existence;

  • as binding force;

  • as applicability;

  • as conformity;

  • as membership;

  • as efficacy.

Nino (1996, p. 118) notes how these “cores of meaning” are not fully autonomous but rather often presented in various combinations—thus contributing to the ambiguity surrounding the use of the concept. Lamond (2014a, p. 113) too observes that, at least in common law jurisdictions,

[t]he language of validity is ordinarily used to make three types of claim about a legal rule: (i) that is it legally effective (has legal force); (ii) that it is a member of this legal system (a valid rule of English Law); and (iii) that is validated by another law (valid under the relevant primary legislation, not ultra vires). There is an important relationship between the three, inasmuch as the standard explanation for why a rule has legal force is that it is a member of the system in question, and the most common basis for a rule being a member of the system is that it is validated by another rule of the system. Validation gives rise to membership, and membership qualifies a rule for normative force. But it is the normative force of rules – their legal effect – that is fundamentally associated with validity (italics original).

I begin by stressing that, as a preliminary methodological matter, the fact that three (or more) properties have “an important relationship” does not warrant the use of one and the same concept for referring to them.2 More importantly, the ambiguity surrounding the use of this concept goes to a deeper, metatheoretical, level. For two very different questions about law have been simultaneously addressed by it, a “theoretical” and a “normative” one (Peczenik 1989, p. 175). That is, in some cases, validity is conceived of as the hallmark—the “specific mode of existence” as we shall see—of law: as Raz (1979, p. 146) puts it,

a rule that is not legally valid is not a legal rule at all. A valid law is a law, and invalid law is not.

Here validity is used in a descriptive manner. In other cases instead, validity means a normative, prescriptive judgment regarding legal norms: to say that a norm is valid is to say that it ought to be followed or applied. A first level of confusion arises when these two different types of discourse—descriptive and prescriptive—are not kept properly distinguished. This leads us to Hans Kelsen.

2.2 The Ambiguity in Kelsen

In what might be one of the most quoted excerpts in modern legal theory, Kelsen (1946, p. 30) affirms:

to say that a norm is valid, is to say that we assume its existence or – what amounts to the same thing – we assume that is has “binding force” for those whose behavior it regulates.

Thus, existence, validity, and binding force of law are synonymous; to wit, they express the same property.3 Or, as Guastini (2016) puts it, there seems to be “conceptual identity” between the three properties. Before looking at the implications of this identification, it is worth noting how, according to Ferrajoli (2016), Kelsen’s semantic ambiguity about validity originates in a higher-level syntactic ambiguity due to his use of the term “norm”. Even though Kelsen (1960) clearly recognizes the fundamental distinction between normative acts—as acts of will that belong to the Sein—and norms, which are of those acts the meaning and belong to the Sollen, he denotes both of them with the same term “norm.” This in turn means that for him (and his followers on this point) it is conceptually impossible to distinguish between the validity of the act that produces the norm and that of the norm itself: hence the equivalence between existence and validity of norms.

What then of the further equivalence with binding force? The passage above is used, among others, by Nino (1978, p. 358) to argue that Kelsen does not hold a purely descriptive concept of validity, but a normative one. In this sense, to predicate the validity of a norm is to make an evaluative judgment about its obligatoriness and justification, that is, to prescribe that the norm ought to be obeyed.4 The problem is that such an evaluative stance runs against the postulate of purity of Kelsen’s theory (Guastini 2016, p. 403). Things are not better if, following Bulygin (1998), we understand Kelsen’s validity as expressing merely a positive legal obligation, for in this case we preserve the purity of the theory at the cost of running into the problem of the necessary extralegal or nonnormative character of the Grundnorm (Guastini 2016, p. 404). In short, either way, the theoretical coherence of Kelsen’s theory is compromised.5

2.3 The Results of the Ambiguity

For the purposes of this chapter, nothing hinges on whether Kelsen’s notion of validity is, after all, descriptive or normative. What matters instead is that in either case, the ambiguous nature of his notion of validity (and of his notion of “norm” before that) yields far-reaching consequences for our understanding of modern legal systems. In particular, the equivalence between existence, validity, and bindingness entails on the conceptual level that
  1. (I)

    all existing legal norms are also always valid and binding;

  2. (II)

    invalid legal norms do not exist and have no binding effect;

  3. (III)

    all that is necessary for a legal norm to be valid and thus binding is the (formal) validity of its source.


These three theses are hard to reconcile with our current legal practices.

(I) implies that validity and binding force become part of the very definiens of “legal norm” so that to talk of a nonbinding legal norm amounts to a contradiction in terms (Peczenik 1989, pp. 175–176). Yet even if we conceive of bindingness in purely legal terms, we know that such contradiction does not hold as a matter of logical necessity (in our modern legal systems at least): just think about the disapplication of domestic norms that are in conflict with European Union legislation. Such norms retain their full validity (provided they had it in the first place) and yet “lose” their bindingness (or force of law).

(II) instead makes it impossible to coherently account for a great number of legal norms (as contained in statutes, secondary legislation, or administrative acts) that might be declared invalid by a court at some point and yet have been part of the legal system—and might have produced legal effects—for a while.6

(III) has the even more paradoxical effect of preventing Kelsen from fully exploring what is likely one of his greatest contributions to legal theory, namely the identification of the static as well as dynamic character of modern legal systems (Ferrajoli 2016).

Kelsen in fact shows us that one of the distinguishing characteristics of law as a normative order is that it regulates its own creation. It does so through the “chain of validity,” which pertains not just to the forms of the acts that produce norms but also to the content of those norms. Yet if we identify the sufficient condition of validity of legal norms with the validity of the acts that produce them, it does not actually matter whether the content of those norms will conflict hierarchically with that of the superior norms.7 In other words, the material validity of a norm—its conformity with higher norms—becomes irrelevant as it is neither a necessary nor a sufficient condition for the existence of the act that produces it (Ferrajoli 2016, pp. 92–93). This leads us to the striking realization that, if we were to take Kelsen at face value on this point, the concept of validity would be ultimately “completely worthless” (Guastini 2016). Is this really the case?

Before we look at the structure of modern legal systems to see whether validity is really a dispensable concept that plays no real function in our understanding of law, we need to clarify what sense of the “existence” of norms is at stake here.

3 Some Remarks on Normative Ontology

As I have mentioned already, the discussion about validity is muddled not only by terminological differences, but also by some deeper conflicts pertaining to the metaphysical nature of norms that are not always brought to the surface of the discussion. In particular, the idea of the “existence” of norms is in many respects puzzling: it amounts to one, if not to the main driver of the naturalistic explanation of law that we know (in its many variants) as legal realism.8 For one thing, norms do not exist in the same way as material objects do as they lack, for instance, mass. In this sense, norms are “mysterious beasts” (Brennan et al. 2013, p. 2) and are more akin to so-called abstract objects, such as mathematical entities and properties (Berto and Plebani 2015). The puzzling metaphysical nature of norms has become relevant again in legal theory since the introduction of linguistic and conceptual analysis and of the key distinction between norms and norm formulations (Narváez Mora 2015).9 Once it was clarified that norms cannot be the norm formulations through which they are expressed—for norm formulations almost always underdetermine norms as their meanings10—the question of what kind of entities norms are does not seem to have a readily available answer any longer.11

As Guastini (2013, pp. 146–147) points out, when we talk about the existence of norms in jurisprudential discourse, we might be alternatively referring to two different questions: (i) what does it mean that a norm exists, in general, and (ii) when does a norm exist in a given legal system, that is, when does a norm belong to it? The first is a metaphysical question that refers to norms in general, while the second is a contingent, positive question whose answer depends on the legal system taken into account—as different legal systems have different criteria as to the membership of their norms. Given space constraints, I will not be able to discuss here all the ontological and metaontological commitments underlying the following remarks,12 but I will nonetheless flesh out a rough account of what it takes for a (legal) norm to exist.

Let us start from the intuition that norms do not make sense as such until they are expressed or formulated: in other words, they do not exist before someone expresses them. As Kelsen (1949, pp. 483–484) holds, there are no normative principles in the physical world, only regularities of behavior and physical laws (such as gravity). The point is that these regularities and physical laws exist—and causally interact with the world—independently of anyone observing and expressing them: they are strongly mind-independent. The same cannot be said of normative principles. A normative principle is such only if it can be potentially acted upon, that is, if it can make a difference, by being taken into account, in the practical reasoning of the agent: and in order to do so, it must be somehow expressed and recognized as such.13 In other words, a normative principle must have some degree of epistemic accessibility if it purports to guide conduct: as Raz (2011) affirms, it must “be in the world.”14 And as expression is an act of will, one could say then that the normative element in any standard or general requirement is necessarily intentional.15

“Expression” seems to be then a necessary, albeit not sufficient, condition for something to be considered normative, for only with its expression does a norm acquire the capacity to interact “causally” with the furniture of the world (Guastini 2013, p. 147).16 We can call this the expressive or epistemic condition of existence of norms. In this first sense, to affirm that “norm N exists” means “someone expressed N” (Guastini 2013, p. 147).17 This might be why many think that norms are a particular kind of abstract object, namely “linguistic entities,” and as such bear a special relationship to language (Moreso and Chilovi 2016; Pino 2016).18 But is this all there is to the existence of norms? Is it enough that someone—anyone—expresses or formulates a deontic requirement and this fact alone grounds its existence qua norm?19

While expression is a necessary condition of the existence of all norms, it is clearly not a sufficient one. I could indeed let you know that “people ought not to listen to music in this city,” which is a syntactically and semantically well-formed deontic requirement. But my utterance could only express a genuine norm in light of something else being the case: for instance, if there is a moral principle that prohibits people from listening to music in this city or if the person uttering the deontic sentence is the sovereign of the system and so forth. Thus, in this second sense, to say that norm “N” exists means “something metaphysically grounds normative principle ‘N.’” We can call this the metaphysical or grounding condition of the existence of norms. This grounding relation obtains always between a norm and a “grounding fact,” which in turn seems to change depending on the type of norm we are considering (von Wright 1963):
  • Moral norms: here there is a preliminary distinction to be drawn depending on whether we are talking about an objectively valid moral norm or a norm of positive morality of a given social group.
    • Critical morality: a norm “N” exists if there is an objective normative principle “N” of morality. Here, in other words, we have the conceptual identity between the truth maker and the proposition that expresses the normative requirement, the problem being instead the existence and epistemic accessibility of such grounding facts (one must be a moral realist, broadly understood).

    • Positive morality: if we consider instead a given social group in a given moment in time, a norm “N” exists if the corresponding normative attitudes are displayed by a “significant proportion” of members of the group (and a significant proportion knows about this) (Brennan et al. 2013, p. 29).

    What is important to point out is that in both cases, the relevant norm “N” exists but is independent from the underlying N-practice—that is, the social practice that sees “N” as its object, or practice of “N”-ing. This is clear in the case of the norms of critical morality, which—postulating the existence of the corresponding objective normative principles—bind irrespectively of the mental states of their addressees, but it is also what, according to Brennan et al. (2013, pp. 58–59), distinguishes the norms of positive morality of a group from its social ones.20

  • Social norms: a social norm “N” exists in a given social group if a significant proportion of members of the group accept “N” and this acceptance is known to a significant proportion of members of the group—so that there is an (at least presumed) N-practice.21

  • Legal norms: a legal norm “N” exists if it belongs to legal system S, for its existence is tantamount to the membership of the system (Pino 2016; Nino 1996). Now what it takes for a norm to belong to system “S” is a contingent question, the answer to which will be different from system to system. The key point for our purposes—and what distinguishes legal norms (qua formal or institutional)22 from moral or social ones (informal)—is that, in this case, the creation and change and eventual demise of norms is regulated by a second set of rules belonging to the system itself (Hart’s secondary rules) that are administered by some type of authority. Law is in this sense reflexive or autonomous. To put it with Kelsen (1946, p. 132), one of its distinguishing features is that it regulates its own creation and application. As such, a legal norm “N” is practice-independent vis-à-vis the corresponding social practice “N” (as much as an equivalent objective moral norm “N1”).23

There seems to be then a key difference between moral and legal norms on the one hand and social ones on the other. The latter, but not the former, are necessarily metaphysically practice-dependent. In other words, a moral or legal norm does not necessarily need a corresponding social practice (not even a presumed one) to exist. This leaves their expression (or formulation) as the only epistemic dimension they might ever “possess,” given that the corresponding social practice might never develop.24 This is clearly shown in the case of legal norms, for example, those that in most countries around the world prohibit the file sharing of copyrighted material. The fact that people do actually share and download illegally copyrighted material on a regular basis—so that the corresponding social practice cannot be said to exist, not even presumably—does not falsify, by itself, the assertion that such legal norms exist (and are valid).

Why is this relevant? According to my analysis, expression is the only ontological (qua epistemic) condition that norms of all kinds share. But for moral and legal norms, given their practice-independence, expression might be and remain their only epistemic display—that is, the only way they might “be in the world” to interact with it. Two conclusions stand out then: first, that for these two kinds of norms, it is even more the case that to affirm that they exist is tantamount to saying that someone has expressed or formulated them25 and that, because of it, the possibility and circumstances of their expression become the real “battleground” in relation to their existence. In the case of norms of critical morality, the question is whether there are such things as objective moral facts (or goods) and, if there are, whether (and how) we can access them. In the case of legal norms instead, the question is whether whoever formulated the norm had the authority to do so and whether this expressive act was carried out within the formal and material requirements imposed by the higher-ranking norms in the system. This brings us back to the concept of validity.

4 Explaining the Validity Vis-à-vis the Existence of Law

Incidentally, the topic of validity is an example of a certain degree of insulation sometimes displayed by Anglo-American jurisprudential scholarship vis-à-vis its continental and Latin American counterparts. Consider again the paradigmatic statement by Raz (1979, p. 146), for whom a rule that is not legally valid is not a legal rule at all. Like many others in Anglo-American scholarship, Raz identifies the validity of law with its existence. Only valid law belongs to the legal system; invalid law does not. In this regard, laws are like stones: “a non-existent stone is not a stone, though we can talk about such stones and describe some of their properties as we can do about invalid rules” (Raz 1979, p. 148). But is an invalid law necessarily not “law”? Intuitively, one can point to all those instances in our current legal systems in which invalid legal acts—like statutes or administrative measures—may still yield normative consequences, at least until the moment when they are annulled or repealed by a court. These acts, albeit invalid, do belong to the legal system—they have some relationship or status within it (Guastini 2013, p. 132).

While the distinction between the existence and the validity of law has come to the fore of Anglo-American scholarship only very recently,26 it has instead been accepted and discussed for quite some time in continental and South American scholarship, following the seminal work of scholars such as Eugenio Bulygin, Luigi Ferrajoli, and Carlos Nino. These scholars have pointed out that Kelsen’s conflation of the existence of a legal norm with its validity conceals what amounts to the key innovation brought about by modern constitutional systems, namely the very possibility of existence of what has been aptly termed by Ferrajoli (2007) “unlawful law,”27 something that would amount to a contradiction in Raz’s (and Kelsen’s) theory.28 Yet this constitutes possibly the most striking feature of modern juridical phenomenology: it points to the fact that there are limits to what even the highest legislative authority in a legal system can do—and these limits are legal ones, as established in some sort of constitutional settlement (usually codified and entrenched but not necessarily). As such, there are some things that as a matter of law the legislator cannot do (Grellette 2010, pp. 26–31): an idea simply unentertainable if we consider the idea of sovereignty as expressed by Bodin (1586) with the idea of potestas legibus soluta and that still nowadays proves hard to reconcile with the principle of parliamentary sovereignty in English constitutional scholarship.29

4.1 Formal and Material Validity

Only by distinguishing between the existence and the validity of law can we truly understand the meaning of the latter—a property that refers to the conformity of both acts and norms to the hierarchically superior norms governing their production. As already mentioned, in its most general understanding, the property of being “valid” indicates the state of being in accordance or conforming with something. The first step toward understanding how “unlawful law” is even conceivable is to recognize that in institutional-normative systems, validity (as conformity) can assume two different dimensions, the formal and the material one. According to Pino (2014, p. 207), the former pertains to

the attainment of the formal/procedural conditions regarding the exercise of law-making power, according to which a certain text can be considered as a legal source. Formal validity usually coincides with a successful “enactment”, and it can be ascertained by means of a factual inquiry about the realization of the relevant law-making procedures (italics mine).

Hence, a bill that is not approved by the Italian Parliament following the prescribed procedure (for instance, a constitutional bill that is approved through the ordinary legislative procedure) is not formally valid, and the Italian Consulta (the Constitutional Court) can quash it (article 136 of the Italian Constitution). Another fitting example is that of a written contract that is lacking one of its essential elements, such as the signature of one of the contracting parties (when this is required by law). Material validity, instead,

obtains when a legal norm is coherent (or at least not conflicting) with the relevant higher-rank legal norms. Plainly, material validity is not a matter of fact, but a matter of interpretation: it depends on the content, on the meaning of the relevant norms. More precisely, it requires interpreting both the norm whose validity is to be ascertained, and the (higher) norms that act as the parameter for the validity of that norm (Pino 2014, p. 208).

Going back to the example of a statute, if the normative content of a formally valid statute violates one of the provisions in the Italian Constitution—for instance, by unduly limiting the freedom of the press to publish news about the current government—then the Consulta will be able again to strike the statute down. The same applies, mutatis mutandis, in the case of contracts: if a contract, for instance, has an illegal purpose (the infamous contract for murder), no judge will hold that contract valid and enforce it. It is important then to stress how formal validity and material validity work on different levels, pertain to different objects, and are a product of different epistemic endeavors.

As Table 1 illustrates, formal validity pertains to normative sources—acts—and requires the attainment or application of certain formal/procedural conditions to validly exercise a power recognized by the law.30 As such, it is ascertainable through an empirical enquiry (whether the bill has received royal assent, for instance).31 Material validity, on the other hand, pertains to norms themselves and requires (at least) the noncontradiction between the content of norms so produced and the relevant, higher, norms. If we look at our current legal practices, this first distinction seems rather uncontroversial. What happens, though, when an act is formally valid but the norm so produced is not materially so?
Table 1

Formal and material validity compared


Formal validity

Material validity

Pertains to





Coherence or noncontradiction

Ascertained by

Empirical enquiry (plus minimal interpretive operation)

Interpretative operation

4.2 Formally Valid but Not Materially So—The Mysterious Status of “Unlawful Law”

We have seen that for several Anglo-American legal theorists who have been following Kelsen on the point, validity is existence and hence validity is bindingness. As has been shown instead by those scholars who warn of the unsatisfying theoretical implications of Kelsen’s thesis (Munzer 1972; Ferrajoli 2007; Grellette 2010; Guastini 2016), this “flattening” of validity with existence leads to a variety of explanatory shortcomings that leave an important amount of legal practice unsatisfactorily unaccounted for. This discussion has taken place mostly at the constitutional level when some sort of “constitutional entrenchment” is present in a given legal system. Interestingly, the distinction between formal and material validity is operatively present even in common law jurisdictions where such constitutional entrenchment is missing—for bylaws, administrative acts, and contracts can all have either formal/procedural or material requirements. Yet its general importance and role as to the ontological status of legal norms in our modern systems has been manifestly undertheorized. Validity in the formal sense only obtains when all the formal requirements on the production of a given act (a contract, a statute, a judicial decision, or a will) are attained by the agent(s) in producing it. Guastini (2013, p. 132) aptly reminds us that, “in lawyers’ talk,” validity means precisely the absence of “vices.”32 But what about the case in which only some of those formal/procedural requirements are attained?

Take the example of a will: a clear formal requirement for a will to be valid is that it is signed (often necessarily in front of some witnesses) by its testator. Now, think about a will whose signature has been forged—it was signed not by its testator but by her nephew, who also (coincidentally) appears to be now the beneficiary of the entire estate. The will looks like a formally valid one to the nonexperts—only an experienced calligrapher would be able to ascertain that it was not actually signed by the supposed testator. It is hard to deny that, until the moment in which a judge declares the will invalid because it lacks one of its essential formal requirements, that will exists—that is, it belongs to the legal system. For one thing, the nephew might have brought some possession claims on the estate on the basis of the forged will. Unless the legitimate heirs to the estate can somehow prove immediately that the will is forged, it is likely that the court will, prima facie, grant the nephew possession of the estate. But even more to the point, the very fact that a court takes the will into consideration and declares it invalid implies in the first place that some type of act that is intelligible as having a juridical meaning (Ferrajoli 2007, p. 528) must have been produced and hence, for this fact, exists for the legal system. This seems also confirmed by the traditional definition of a nullity in law as an act that must be treated as if it had never existed or taken place. This logically implies that for a certain amount of time, and precisely from the moment t0 of creation to the moment t1 of declaration of nullity by the judge, an act juridically meaningful existed for the legal system.33

When is it that a will does not even exist for the law then? If we observe our juridical practices, that would be when such a supposed will lacks even some of those foundational requirements that allow for its very recognizability as a juridical (type-)act—for instance, a will that lacks the written form. If someone were to tell me in person that she wants to leave her estate to her older child but not to the younger, reckless one, I would never take that to be a speech-act that has any legal relevance whatsoever. Even though my friend is genuinely moved by the desire to make her intentions regarding her estate manifest, no one in her sound mind would take such a speech-act as constituting an enforceable legal act. In other words, the requirement of the written form is so fundamental when it comes to the concept of a will that in its absence there is really nothing we can talk about from the legal point of view.34,35

4.3 The Artificial Ontology of Modern Law

The discussion above points to a key aspect of the difference between juridical existence and nonexistence in modern legal systems. For the vast majority of legal acts (contracts, statutes, administrative and judicial decisions, wills, etc.), most of their necessary requirements to exist in the legal system will be knowable and ascertainable by laypeople and not by legal officials only.36 Anticipating my main claim in the next section, the point is that we can shed light on the complex ontological status of modern law only if we bring to the surface the underlying epistemic issues. Whether a norm belongs prima facie to the legal system or not must be, in normal situations, tendentially knowable and ascertainable by rational and linguistically competent agents in the system. This amounts to a necessary condition—given the practice-independence of legal norms—for law to be tendentially objective and thus able to guide the conduct of its addressees.37

Whether a feasible inheritance intention has been expressed in written form is something that any person of sound mind can ascertain, and so is whether the document has been signed or not.38 The same applies, mutatis mutandis, to statutes: if the British Prime Minister were to publish a Facebook post with a list of general rules addressed to the public and claiming these were new statutory law, one would take her as either
  1. a)

    reporting a very recent legislative development no one has ever heard of; or

  2. b)

    joking; or

  3. c)

    being delusional.


The point is that no one would think that her speech-act could possibly constitute a statute or legislative provision, and if someone did, well that someone would arguably lack the very concept of a statute.39 Whether a bill has been correctly approved by a given parliamentary commission or whether the signature at the bottom of it by the monarch or the president of the republic is present or not is again an empirical question that can in principle be answered by a great many people in our modern legal systems.

In short, the distinction between (mere) existence and formal validity of an act is premised on the observation that existence requires the application of some of the essential formal norms on the production which make the ensuing (product-)act juridically intelligible to every rational and linguistically competent agent in the system.40 If that act also contains a norm formulation that expresses a deontic sentence, we can say then that a new norm that belongs prima facie to the system has been created. An act that exists for the legal system yields a juridical meaning that is recognizable as such by laypeople and officials alike. It might be lacking some formal elements that are necessary to be considered formally valid, or its content might contradict some higher norm and thus be materially invalid, but at this stage it is still the case that upon the mere attainment of some formal/procedural requirements, this act has the potential to produce a change in the normative landscape of its addressees.41 It is prima facie a source of law.

Table 2 below illustrates the complex ontology of modern law,42 which is the result of the fundamental distinctions between acts (as containing norm formulations) and norms (as the meaning of those acts) and between their respective existence and validity. According to this model, validity can be predicated of two different entities: of acts (formal validity) and of norms (material validity). The former is (predominantly) a matter of empirical conformity, the latter of interpretive coherence or noncontradiction (Ferrajoli 2016, p. 85). This means that every time someone refers to law as being valid with no further specifications, she usually means the combination of both—a statute is valid only insofar as it is so both formally and materially. This consideration allows us also to make intelligible the status of “unlawful law”: that is, all those statutes, regulations, bylaws, judicial decisions, and so forth that albeit extant within the system might be actually invalid. As such, the existence of a legal norm (and of the act by which it is expressed) does not imply its validity. Existence is a necessary but not sufficient condition of validity.
Table 2

The artificial ontology of modern law



Formal validity

Material validity

Pertains to

Acts (sources)

Norms (contents)

Acts (sources)

Norms (contents)


Attainment of at least some essential elements


Coherence or noncontradiction

Ascertainable by

Every rational agent (in normal situations)

Juridical operators (officials, lawyers, etc.)


Ascertained through

Empirical inquiry (plus minimal interpretive operation)

Empirical enquiry (plus minimal interpretive operation)

Interpretative operation

Normative effect

Force of law

Force of law (enforceability if there are no material norms curtailing the contents)

Force of law and enforceability

Two remarks are in place at the end of this section. Interestingly (and perhaps ironically, given that this chapter is premised on a critical reading of this author), each of the remarks seems to vindicate an important intuition by Kelsen. First, the analysis shows in what sense validity can be conceived of as the specific mode of existence of legal norms vis-à-vis social and moral ones. Once validity is “released” from its normativist declination, we are able to appreciate that “valid” is a second and “artificial” ontological status that legal norms—qua formal or institutional—can entertain beyond that of “existence.” In this sense, validity is a property that originates necessarily from the institutional character of a system of norms43—it is a relationship between norms and the acts that create them (Pino 2016, p. 106). This is the way in which law regulates its own creation, which for Kelsen is precisely one of the distinguishing features of law in respect of other normative systems.

The second and generally undertheorized44 point that our analysis underscores is that, in our modern legal systems, a norm has force of law already with the mere existence of the act that produces it and not just with its (full) validity. Granted, it is necessarily a prima facie force, or, to put it more aptly, there is a presumption of force of any norm produced by an act that exists within the system.45 That presumptions only stands until a norm is scrutinized by a court as the court will not enforce invalid norms, only valid ones.46 But it appears as an undeniable descriptive truth about our legal practices that, until an invalid act is brought before a relevant official for the authoritative declaration of its invalidity, such an invalid act might nonetheless produce normative effects like the equivalent valid one.47 Hence, pace Kelsen, validity and force of law are independent properties that do not entertain any relationship of entailment. An invalid norm, as long as the act producing it exists for the system, can exert force of law, and if it goes unchallenged, the normative consequences thus produced might stand (Grellette 2010, p. 28). This also explains why modern legal systems allow claimants who wish to challenge the validity of a given act to seek “interim injunctions”—measures with which a court might crystallize the (normative) status quo while the legal proceedings unfold and until a decision is taken on the validity of the act in question. This central feature of our current legal practices would make no sense if invalid acts were to be considered nonexistent (and thus unable to produce any effect whatsoever).

5 Validity and Objectivity: On the Epistemic Dimension of Law

This chapter started off with the demonstration of how one of the greatest jurisprudential minds has fundamentally misunderstood the relationship between the existence, validity, and bindingness of law. This has subsequently led many more in the literature astray. I have also shown that the consequences of this mistake cannot be possibly underestimated. For validity becomes either a redundant concept that bears no real explanatory capacity vis-à-vis our modern legal practices, leaving a great deal of them unaccounted for, or it must be understood as the expression of an evaluative judgment that implies some form of “ethical legalism” or “ideological positivism” (Guastini 2016). Instead, once properly distinguished from “existence” and conceived of as part of a complex ontological paradigm, the concept of validity is a key element to shed light on the inner structure and workings of modern legal systems (vis-à-vis other normative practices).

In this sense, law is widely held to be, as an institutional-normative practice, a means of social organization and control: not the only one, and not necessarily the best one, but a specific one (Kelsen 1941). The specificity of law lies in subjecting human conduct to the guidance of norms (Fuller 1969), in this way providing its addressees with reasons for action. Green (1998, p. 121) and Gardner (2012, pp. 205–211) have claimed that law might be a modal and not a functional kind: it would be defined and identified not by what it does (for other normative practices such as morality or custom perform the same function) but by how it does it. On this point, I stand by Ehrenberg (2009, p. 97) instead when he affirms that

[w]hat is unique may be neither its form (modality) nor its function but instead the particular way in which the two are joined.48

The “particular way,” as I noted above, is to use norms to guide people’s conduct. This practice might originate from the spontaneous dynamics of a given social group, but once it reaches a certain dimension and level of complexity, the set of (primary) norms as established within those spontaneous practices will no longer be able to effectively and satisfactorily work as a means of group organization and control. Here, of course, I am just picturing in broad strokes Hart’s (2012) reconstruction of the evolution from premodern legal systems to modern ones through the union of primary and secondary rules, which I fully endorse.49 What is important for our purposes is that the evolution from a set of (social) primary norms to a system of primary and secondary norms entails the ontological independence of those primary norms from the corresponding practices, of which these latter are the most obvious epistemic manifestation in spontaneous social settings. In other words, insofar as social norms are informal and entail the corresponding practices, members of the group have a reliable and conspicuous epistemic source to find out what they ought to do. But when instead the creation of new norms becomes an institutionalized process in the hands of a particular subject (i.e., group leaders or an assembly), a new norm “N” can be created and thus exist even if no one (in the most extreme case) in the social group displays the corresponding “N”-normative attitudes.50 Norms thus created are fully posited precisely because there is no necessary relation with the corresponding practices. This requires in turn the centralization of coercive enforcement within the group to “cope” with physiological norm breaching by some members of the group.51 The positivization of norm creation and the centralization of coercive power appear to be then inextricably related to one another. As Postema (2011, p. 307) puts it:

[w]ith law comes the institutionalization and centralization of governance or exercise of political power.

The result of these processes is twofold. On the one hand, the positivization of norm creation requires the social group to introduce new norms (and change existing ones) in a way that is epistemically accessible through the institutionalized exercise of authority (see Westerman in this volume). In other words, legal norms thus created are ontologically grounded in forms or processes that are empirically observable. This is famously captured by the Hobbesian maxim auctoritas, non veritas, facit legem.52 On the other hand, at this first stage, whatever content can be poured into the form of the law; to wit, at this stage, law merely channels the exercise of political power, which remains (potentially) unfettered (Ferrajoli 2016). It is only then with a second and artificial ontological dimension—validity—that law itself regulates its own creation, so that there are legal limits to what the authority can and cannot do. And it is precisely with this limitation of law by law that also comes logically the possibility of “unlawful law,” which has proved so hard to reconcile with positivist theories of law (Grellette 2010, pp. 28–37).

Perhaps what many positivists have struggled with is the fact that legal norms are produced through an empirical act of creation but also entertain a normative relationship of validation with superior norms (Lamond 2014a; Raz 1979, p. 150). This presupposes Kelsen’s binary typology of normative systems that are either “dynamic” (where validity depends on the procedure of creation of that norm) or “static” (where a norm is logically derived from other norms in the system).53 As should be clear by now, our modern legal systems—no matter whether based on the civil or common law tradition—are static and dynamic at the same time (Ferrajoli 2007; Beltrán and Ratti 2010): there are two logically independent modes of validation of legal norms—one source based (or formal), the other content based—that are at work simultaneously because of two expedients. First, this is because the source-based criterion of validation is already by itself sufficient to give rise to the force of law for the norms thus produced, at least prima facie. Second, while the ascertainment of the fundamental elements of legal acts (such as statutes, contracts, wills, etc.) that constitute their recognizability as sources of law can be carried out by every rational and linguistically competent member of a group, the authoritative decision of “invalidation” for the norms thus produced is reserved for a closed group of specialized agents within the system. In other words, unless an act (source) is challenged before the authority entitled to pronounce on its validity, that potentially less-than-valid act might yield the same normative effects as a valid one.54

Episodes like these happen regularly in our legal practices. And taken individually, each of these instances of “invalid law” that still produce normative effects constitutes a shortcoming of the system, a glitch. But the very possibility of “unlawful law” (and of its occurrences) must be understood as the price we have to pay to have an institutional-normative system that regulates its own change while being (tendentially) objective and thus allowing for predictability and certainty.55 The always potential and irreducible tension between the two independent processes of validation of legal norms is resolved by a default rule according to which the source based prevails over the content based, unless there is a supervening authoritative act (by a court) that sanctions the invalidity of the act in question. To put this in Kelsenian terminology, the static character of the system is partially sacrificed to its dynamic one (Ferrajoli 2007, 2016; MacCormick 2008, p. 162). Its internal coherence—which would require the logical impossibility of “unlawful” law—is sacrificed toward its certainty and predictability, two features that are necessary to fulfill its guiding function of subjecting human conduct to the guidance of rules (Peczenik 1989, p. 174).

Could this be any different?56 If “force of law” were only a predicate of the full validity of a norm and not just of the existence of its source, a normative system could fulfill its guiding function only through the following two “routes.” Either it would be possible for officials to express an ex ante authoritative judgment of validity, both formal and material, for every (token-)act that purports to produce normative effects as recognized by the law, or such authoritative judgments of validity would not have to be reserved to a restricted group of specialized agents (the courts), but would instead have to be diffused, in the sense that every agent in the system could authoritatively establish what is valid and thus binding law. The unavoidable problems these two possibilities should be apparent. As to the first possibility, it is not clear how a state of affairs in which officials perform an ex ante validity control on each and every act of law creation would be achievable in our large societies, whereas in the second case, law would lose even that (minimal) type of objectivity, that is, intersubjectivity, and become an unavoidably subjective practice: one in which everyone in the system is able to say what the law is—even the lunatic (Jori 2010). I guess there is no need to stress how such a system—if we can talk indeed of a “system”—would hardly achieve any action-guiding function at all as no one could reliably form any expectation vis-à-vis the behavior of others.

6 Conclusion

Once released from its normative (mis)understanding, the concept of validity assumes a completely different explanatory role vis-à-vis our modern legal practices. As an artificial ontological status that legal norms can possess beyond that of (mere) existence, it allows law to regulate itself (rather than being just the form of exercise of political power), thus creating the logical space for the existence of the only apparently contradictory “unlawful law.” This is a crucial point toward shedding light on the relationship between the ontological and the epistemic dimensions of objectivity of modern legal practices. In this sense, due to the ontological practice-independence of legal norms, the institutionalized creation of law must be epistemically accessible for objectivity—which Kramer (2007, p. 46) has termed “transindividual discernibility”—to obtain. In other words, the fact that the existence of a source of law depends on formal, empirical features that are in the majority of situations ascertainable by every rational agent in the system (like the written form of a deed or the fact that it has been signed and so forth) grounds the status of law as a tendentially objective (that is, intersubjective) practice. Importantly, this objectivity is grounded not only on the practice of the officials of the system, as is often held, but on the practice of laypeople as well.57 This is the epistemic importance of easily identifiable conventional rules that Hart (2012, p. 134) stresses in the Concept as it is this characteristic that allows the law to fulfill its general action-guiding (and thus coordinating) function. A given deed is easily recognizable as a source of law (that is, as a token of the juridical type to which the system assigns the production of certain normative consequences) due to the occurrence of some formal characteristics. A statute approved according to the required voting procedures and promulgated in the official gazette is law, whereas a list of norms published on Facebook is not law (at least for the time being).58 It is this kind of epistemic objectivity that, at a minimum, grounds the rule of law.59


  1. 1.

    According to Sartor (2000, 2008) validity should instead be understood as expressing a “doxastic obligation”: that rule R is valid means that we ought to accept rule R in our legal reasoning.

  2. 2.

    See Ferrajoli (2007, pp. 56–57) on why syntactic clarity is a matter of the utmost importance in any analytical reconstruction of law.

  3. 3.

    Kelsen is not alone in confusing validity for something else: while natural lawyers “tend[ed] to identify legal validity with justice”, so that “a norm is ‘properly’ valid only if it is coherent with certain objective values,” legal realists identify validity with efficacy, so that valid norms are only those “implemented in society” (Pino 1999, p. 535). A notable exception is Finnis (2011, pp. 25–29) who, following Aquinas, recognizes the distinction between the existence of positive law—depending on a valid enactment—and its full validity that depends on its “derivation” from natural law.

  4. 4.

    It is interesting to note that when validity is predicated of moral and social norms, it seems to be meant exclusively in the normative sense—we claim that a moral or social norm is valid when we are saying that it ought to be obeyed. For if we were instead to use validity to address the ontological dimension, the concept would be simply redundant: there is no ontological status other than that of existence for moral and social norms. Anne Ruth Mackor points to me that German terminology seems to warrant my contention, because even though we cannot say of a social norm that it is Gültig (since it is not enacted in accordance with certain criteria), we can say nonetheless that is has Geltung, which means that it exists and it has binding force. This might help to explain the widespread flattening of the ontological question onto the normative one in legal discourse, and the resulting equation between validity (in the sense of existence) and bindingness.

  5. 5.

    See Nino (1996, p. 119). MacCormick (2008, p. 162) states that bindingness is not validity, but “one of the consequences” of it.

  6. 6.

    See Kelsen’s (1992, pp. 72–73) strained attempt to account for “rechtswidriges Recht” in terms of the contemporary validity of “alternative provisions” (both the valid and the invalid one). In this sense legal norms would always have an implicit alternative clause: see the rebuttal of this idea in Guastini (2016).

  7. 7.

    For the criticism that a purely source-based criterion can never help identify by itself “individual legal norms” see Priel (2011). As we shall see, legal norms are actually always the product of the interpretation of the act adopted according to the rules on its production.

  8. 8.

    Most legal realists do not dispute that something we ought to call “law” exists in fact—what they contend is that this institutional practice we call “law” is made (also) of metaphysical entities such as norms. Rather, law is exclusively made of human behavior in their view, and as such it is amenable to empirical observation.

  9. 9.

    For a comprehensive and insightful discussion of the distinction and its ramifications see Pino (2016).

  10. 10.

    Narváez Mora (2015) claims that legal norms “are not entities of any kind” and that the necessity to look for the ontological “substance” of norms is what inevitably leads us into the rule-following paradox. I address this worry in my forthcoming monograph on the distinction between creation and application of law (Sandro 2019, forthcoming).

  11. 11.

    This is a question that belongs to normative ontology in general, and not just to legal theory. I use the definition of normative ontology to distinguish the approach outlined here from the one that pertains to the social ontology of law, on which see the recent debate in Rechtstheorie with contributions by Bernal, Canale, Ekins, and Tuzet. In particular, using Tuzet’s (2014) terminology, here I am chiefly concerned with “entity-ontology,” while the social ontology approach deals with “process-ontology.”

  12. 12.

    As such a great deal of the analysis that follows is based on the nonreductive ontological account of norms put forward by Brennan et al. (2013). While I do not necessarily subscribe to every element of their proposal (especially when it comes to the analysis of legal norms), it does constitute the most comprehensive account of normative ontology currently available in the literature—and it does have a considerable edge over reductive accounts (like the ones that equate social norms with the underlying practices). For a more general discussion of metaontology I refer the reader to the formidable introduction to the subject by Berto and Plebani (2015). For an application of this line of enquiry to law see Narváez Mora (2015) and Moreso and Chilovi (2016).

  13. 13.

    I prefer the use of the term “expression” over “formulation” (the one favored by Guastini) given that sometimes norms can remain “implicit,” that is not explicitly verbally formulated. Yet even implicit norms must be somehow expressed (through behavior, with reactive attitudes for instance) in order to acquire the capacity to figure in practical reasoning. I owe this remark to Sebastián Figueroa Rubio and Anne Ruth Mackor.

  14. 14.

    Cf Brennan et al. (2013, p. 31). On how post-modern law is progressively losing its previous mode of “being in the world” see Walker (2009).

  15. 15.

    For Gaus (2014) norms are “human [normative] artifacts.” Cf also Gardner (2012, pp. 59–65, 85–86).

  16. 16.

    Gardner (2012, p. 86) seems to limit this necessary positive character of norms to legal ones, but I do not see any reason for such limitation. See also the quote by Samuel Pufendorf about the existence of “moral entities” in Westerman’s contribution to this volume.

  17. 17.

    This understanding of ontological parlance in arithmetical terms seems to be similar to the one by Moreso and Chilovi (2016).

  18. 18.

    Cf Narváez Mora (2015, pp. 43–44).

  19. 19.

    For an introduction to metaphysical grounding see Tahko and Lowe (2015).

  20. 20.

    The difference between norms of positive morality and social ones lies in the fact that the former “are constituted by [normative] attitudes that are necessarily practice-independent,” while the latter are “necessarily practice-dependent.” In this respect it is worth pointing out that for these authors social norms only entail presumed corresponding social practices (p. 76) and not actual ones, thus allowing for the possibility of “universal error” (Gaus 2014) on the existence of such corresponding practices. Now while I agree with Brennan et al. and with many others (see for instance Perry 2015, p. 284 fn.3) that to identify social norms with the corresponding practices amounts to a serious category mistake (to the extent that it is not the case that the existence of a practice is ever sufficient to generate a social norm), I am not convinced that a corresponding social practice needs only be presumed (and not actually obtain) for a social norm to exist. For the purposes of this chapter however nothing fundamentally hinges on this point.

  21. 21.

    Here (as well as before with norms of positive morality) I leave it open whether this knowledge condition requires common or merely mutual knowledge amongst members of the group. For an insightful discussion about the acceptance of social rules see e.g. Perry (2015).

  22. 22.

    Institutionalized normative orders—such as those of a company, or of an association—are to be considered in this respect more similar to law than moral and social orders, as it makes indeed perfect sense to speak of a “valid norm of that association” if by this we refer to the (institutional) fact grounding the creation of such a norm (MacCormick 2008, p. 160).

  23. 23.

    While legal norms are practice-independent when considered atomistically vis-à-vis their corresponding N-practices, the legal system as a whole is not so—it is dependent on a different, (general) S-practice of recognizing the authority of the system and on its acceptance by the majority of members of the group. So there seems to be always a social (conventional?) fact that ultimately grounds legal norms and that is not present for objective moral norms, which marks once again the difference between law and (critical) morality. Brennan et al. (2013, p. 49) express this difference (in what seems to me a more convoluted way) by claiming that formal norms (of which legal norms are a subset) need only involve de re normative attitudes, and not de dicto ones. See furthermore on the concept of “systemic acceptance” Lamond (2014b). This remark seems also to map onto the distinction drawn by Canale (2014, p. 307) between the social practice which originates a legal system and that must be characterized by “collective intentionality” and the subsequent social practice that “warrants the existence of a legal system over time” but need not be characterized by any such “joint action.”

  24. 24.

    We might see epistemic manifestation in the potential judicial decision that applies and enforces the norm, but this is also a manifestation of the underlying S-practice underpinning the authority of the system.

  25. 25.

    Someone with (normative and not just epistemic) authority, that is. This signals another difference with social and positive morality norms, namely that in this latter case, given their practice-independence, it does not really matter who formulates them (formulation works only as a “proxy” for the corresponding practice). The same does not clearly apply to law, as we shall see shortly.

  26. 26.

    This is mainly due to a series of recent contributions by Matthew Grellette. His work, although commendable, cannot “rescue” Anglo-American jurisprudence from the charge of insularity: while Grellette presents the distinction between the existence and the validity of law as “unexplored” (Grellette 2010, p. 38), this distinction has been put forward and discussed in Continental and Latin-American jurisprudence for over 30 years. This does not seem to be due to a lack of resources in English—as Pino’s (1999, p. 535) article in Law and Philosophy discusses Ferrajoli’s idea that the “mere enactment [of a legal norm] does not coincide with validity.” Similarly, there is no reference to Munzer’s important book Legal Validity (1972) where, almost 40 years before Grellette’s work, Munzer clearly distinguished not only between formal and material validity (p. 25), but also between the existence and the validity of legal norms (pp. 37–43).

  27. 27.

    This is my translation (Sandro 2011) of the original Italian expression “diritto illegittimo.” For a discussion of the concept of “unlawful law” see Grellette (2010, p. 36).

  28. 28.

    Grellette (2010, p. 32) correctly recognizes that early positivists like Austin or Hart took as paradigmatic cases legal systems whose criteria for validity were based purely on conventional practice and hence it would be uncharitable to make too much of their failure to distinguish between existence and validity. The same cannot be said of later positivists, and especially of Kelsen.

  29. 29.

    Think about the discussion which ensued after recent decisions such as R (Jackson) v Attorney General [2005] UKHL 56 and AXA General Insurance v Lord Advocate [2011] UKSC 46.

  30. 30.

    By including both formal and procedural conditions in the definition of formal validity I intend to account for a common ambiguity in legal language between “act” as the process of production of it (i.e. the legislative process) and “act” as the product of such process (i.e. the statute): see Ferrajoli (2007, pp. 494–498).

  31. 31.

    As has been noted, even here there is some degree of interpretive activity, as the formal norms on production will be used to verify the correct procedure and form of the law-making act (Ferrajoli 2007, pp. 524, 577). This remark points to the more general idea that norms, as meanings, are always and only the product of interpretation, an idea that is expressed, for instance, by many legal realists of the Genoa School (following Giovanni Tarello) and, more recently by Pino (2016, pp. 24–26). In this sense it would be more correct to claim that norms would not exist before interpretation. This thesis, taken seriously, has far-reaching consequences for the (in)determinacy of law thesis that cannot be discussed here: I only want to note that when it comes to the attainment of formal or procedural norms, this interpretative activity can be considered minimal and predominantly linguistic.

  32. 32.

    See also Beltrán and Ratti (2010, p. 605) for the claim that “for a rule to be legally valid it must have been produced in a ‘legally impeccable way’, or is a logical consequence of some impeccably produced rule, and is in any case compatible with superior rules.”

  33. 33.

    Cf Pino (2016, pp. 24–25). For the purposes of the present chapter, I am leaving aside the discussion regarding the effects of such acts for the system (that is, the difference between nullity and annullability/voidability).

  34. 34.

    Some civil law expert could rebut: what about nuncupative wills though? It is indeed true that a handful of jurisdictions around the world recognize the existence of oral wills, although they do so only in extreme circumstances where there is no possibility to redact a document following certain formalities. This is the “pure” nuncupative will, that is an oral will whose existence does not depend of any formality whatsoever. It is limited to soldiers at the front and seamen at sea when the death of one of them is imminent. The fact that in these situations no formal requirements whatsoever are requested by the law for the existence and hence validity of the will can be explained by the greater moral bond that arguably develops between human beings before a situation of great peril or danger—like a war or a tempest at sea—and thus by the fact that in these situations one can expect the recipient of the oral will to report it verbatim to the authorities. Alas, the possibility of a completely invented will being reported is still looming here, and that is why some commentators argue that in these cases it would still be preferable to leave the distribution of the estate of the deceased according to the rules of inheritance in the given jurisdiction (see Prascina 2009).

  35. 35.

    It is also true that the oral form was indeed accepted in Roman law. But this did not mean that someone could just report someone’s words claiming they were his testamentary dispositions; rather, there were always some procedural requirements—the oral statement, for instance, had to be given in front of a considerable number of witnesses (usually seven) none of whom was going to become a beneficiary (Prascina 2009). What we have experienced is precisely the evolution of the institution of the will from the oral form to the written one. The reasons for this evolution are the same (i.e. epistemic) that underpin the key relevance of understanding validity as a different status from existence in our legal practices. Thus it is safe to say that barring truly exceptional situations, any existing will must be written. If the ordinary fellow tells a group of friends about his testamentary intentions and considers that speech-act to be his “will” (and so to have normative effects once he passes away), not only does he not know the law and fails to produce any normative effect regarding the distribution of his assets—more importantly, one could wonder whether his concept of “will” is not defective altogether. True, the friend’s speech-act could have several other effects—for example, that of prompting a change in her younger child when he is told about his mother’s intentions—but that does change the fact that it cannot by any means purport to have the effects that the legal system acknowledges when a valid token of the (type-)act “will” is performed.

  36. 36.

    The distinguishing line between requirements that are necessary and sufficient for an act to be existent, and those that are necessary for its formal validity, is a thin one, and seems to resist precise theoretical treatment.

  37. 37.

    Cf Brennan et al. (2013, p. 10). See also Westerman in this volume: her distinction between type- and token-validity maps onto the one between existence and full validity (albeit from a different viewpoint).

  38. 38.

    Whereas the question of whether the signature at the bottom of a specific will has been forged or not is something that only an expert might be able to verify (provided that the forgery is apt, of course).

  39. 39.

    At the very least, a statute is a set of general norms addressed to the public and produced by a prima facie legitimate authority according to some given procedure, be it a monarch, a president, a parliamentary assembly, and so forth.

  40. 40.

    Carpentier’s contribution in this volume is another very laudable attempt to dispel the ambiguity surrounding the use of the concept of legal validity in legal discourse. While he precisely identifies the distinction between the membership (within a system) of a legal norm as its existence and conformity to “higher-ranking norms” as validity, he still terms both of these properties as (two different types of) “validity,” hence somewhat carrying on the ambiguity. I must also confess that I am somewhat not entirely convinced by his claim that conformity with criteria of membership (i.e. existence) is “not a normative, but a conceptual kind of conformity” (original italics). For I do not see any qualitative difference between the test for existence and the test for validity of legal norms in modern constitutional systems (although the former is logically prior to the latter).

  41. 41.

    Cf Gardner (2012, p. 61).

  42. 42.

    Cf Pino (2016, p. 42).

  43. 43.

    On the institutionality of law the locus classicus is MacCormick (2008); for a very recent take see Ehrenberg (2016).

  44. 44.

    A notable exception is Ferrajoli (2007, p. 528).

  45. 45.

    To use Westerman’s terminology in this volume, this would be the function of type-validity (as opposed to token-validity).

  46. 46.

    In reality, it is perfectly possible for a court or tribunal to make an error of law and enforce an invalid norm. If that mistake of law is not rectified by appellate courts, we might well have a change in the law. But the overall objectivity of the system allows us to properly recognize these occurrences for what they are, mistakes of law (at least at first).

  47. 47.

    This to the point that, after a certain period of time has passed, such effects might be “stabilized” by the legal system itself in order to endure certainty and protect the bona fide expectations of third parties: cf MacCormick (2008, p. 162).

  48. 48.

    Later in the piece Ehrenberg defines law as a “social kind.”

  49. 49.

    “Rule” and “norm” are synonymous for the purpose of this chapter. On the most appropriate way to read Hart’s normative “genealogy” see Postema (2011, pp. 306–307). My conclusions are also convergent with Westerman’s account of validity as reputation in this volume. In particular see her considerations about the functional advantages entertained by formal institutional orders vis-à-vis informal ones.

  50. 50.

    As I said, this does not exclude that the social group must somehow accept the normative-institutional system in the first place. Without that basic systemic acceptance, there would be no possibility for norms to be detached from their corresponding practices.

  51. 51.

    In the absence of spontaneous reactive attitudes that come when the practice is there instead. But if the majority fails to comply with the norms coming from the power-holder(s) all the time, one can reasonably doubt whether a legal system is in place at all.

  52. 52.

    The maxim appears for the first time in the 1670 Latin translation of Leviathan (1651): Hobbes (1839).

  53. 53.

    Kelsen’s use of “static” seems very different from Hart’s (2012, pp. 92–93).

  54. 54.

    Contra Kirste, in this volume, according to whom only valid law has the “force of law.” As I have shown in this chapter, this claim seems hard to reconcile with our current existing legal practices.

  55. 55.

    Granted, a system could not fulfill any guiding function if the instances of “unlawful law” were to go beyond a certain, “manageable,” threshold.

  56. 56.

    See for a similar discussion Guastini (2013, pp. 131–135).

  57. 57.

    “Whenever such a rule of recognition is accepted, both private persons and officials are provided with authoritative criteria for identifying primary rules of obligation […] The existence of this [simple form of] rule of recognition will be manifest in the general practice, on the part of the officials or private persons, of identifying the rules by this criterion” (Hart 2012, pp. 100–101). It has been noted already for quite some time that Hart is ambivalent in the Concept as to whether the rule of recognition is grounded on the practice of officials only, or on that of laypeople as well. I submit here that this ambiguity is due precisely to Hart’s lack of distinction between the existence and the validity of law. Once such a distinction is in place, we can understand Hart’s claims as describing not a single, but two rules of recognition, one that provides the criteria for the existence of law (this is sometimes called the “rule of identification”), and the other that provides the criteria for its validity.

  58. 58.

    See again the important contribution by Westerman in this volume. Granted, even the existence and not just the validity of some types of formal acts will be too complex to be ascertained by the nonspecialized members of society; but this technical, more restricted class of acts will still be conceivable only if the general practice of recognizing law’s existence holds for the majority of acts in the system.

  59. 59.

    For an extensive discussion of the relationship between objectivity and the rule of law see Kramer (2007).



Many thanks to all the participants to the “Public law and public trust” seminar in Groningen from which this contribution originates and in particular to Mathieu Carpentier, Jaap Hage, and Miodrag Jovanović for challenging discussions on the topic of validity. My utmost gratitude goes also to Samuele Chilovi, Sebastián Figueroa Rubio, Riccardo Guastini, Nicolas Kang-Riou, Anne Ruth Mackor, and Anastasia Shesterinina, who provided me with extensive comments on previous drafts. The usual disclaimer applies. Finally, this chapter is dedicated to Pauline Westerman: per aspera ad astra.


  1. Beltrán, J.F., and G.B. Ratti. 2010. Validity and Defeasibility in the Legal Domain. Law and Philosophy 29: 601.CrossRefGoogle Scholar
  2. Berto, F., and M. Plebani. 2015. Ontology and Metaontology: A Contemporary Guide. London: Bloomsbury Academic.Google Scholar
  3. Bodin, J. 1586. De Republica libri sex. 1st ed. Paris: Apud Iacobum Du-Puys sub signo Samaritanae.Google Scholar
  4. Brennan, G., et al., eds. 2013. Explaining Norms. Oxford: Oxford University Press.Google Scholar
  5. Bulygin, E. 1998. An Antinomy in Kelsen’s Pure Theory of Law. In Normativity and Norms: Critical Perspectives on Kelsenian Themes, ed. S.L. Paulson and B. Litschewski Paulson. Oxford: Clarendon Press.Google Scholar
  6. Canale, D. 2014. Is Law Grounded in Joint Action? Rechtstheorie 45: 289.CrossRefGoogle Scholar
  7. Ehrenberg, K.M. 2009. Defending the Possibility of a Neutral Functional Theory of Law. Oxford Journal of Legal Studies 29: 91.CrossRefGoogle Scholar
  8. ———. 2016. The Functions of Law. Oxford: Oxford University Press.CrossRefGoogle Scholar
  9. Ferrajoli, L. 2007. Principia Iuris: Teoria del Diritto e Della Democrazia. Vol. 1. Bari: Laterza.Google Scholar
  10. ———. 2016. La logica del diritto: dieci aporie nell’opera di Hans Kelsen. Bari: GLF Laterza.Google Scholar
  11. Finnis, J. 2011. Natural Law and Natural Rights. 2nd ed. Oxford: Oxford University Press.Google Scholar
  12. Fuller, L.L. 1969. The Morality of Law. 2nd ed. New Haven: Yale University Press.Google Scholar
  13. Gardner, J. 2012. Law as a Leap of Faith: Essays on Law in General. Oxford: Oxford University Press.CrossRefGoogle Scholar
  14. Gaus, G. 2014. Review of “Explaining Norms”. Notre Dame Philosophical Reviews (accessed 28/08/2017).Google Scholar
  15. Green, L. 1998. The Functions of Law. Cogito 12: 117.CrossRefGoogle Scholar
  16. Grellette, M. 2010. Legal Positivism and the Separation of Existence and Validity. Ratio Juris 23: 22.CrossRefGoogle Scholar
  17. Guastini, R. 2013. Distinguendo Ancora. Madrid: Marcial Pons.Google Scholar
  18. ———. 2016. Kelsen on Validity (Once More). Ratio Juris.Google Scholar
  19. Hart, H.L.A. 2012. The Concept of Law. 3rd ed. Oxford: Clarendon Press.CrossRefGoogle Scholar
  20. Hobbes, T. 1839. In Opera Philosophica Quae Latine Scripsit Omnia, ed. William Molesworth, vol. 3. London: Joannem Bohn.Google Scholar
  21. Jori, M. 2010. Del Diritto Inesistente: Saggio di Metagiurisprudenza Descrittiva. Pisa: ETS.Google Scholar
  22. Kelsen, H. 1941. The Law as a Specific Social Technique. The University of Chicago Law Review 9: 75.CrossRefGoogle Scholar
  23. ———. 1946. General Theory of Law and State. Cambridge, MA: Harvard UP.Google Scholar
  24. ———. 1949. The Natural-Law Doctrine Before the Tribunal of Science. The Western Political Quarterly 2: 481.CrossRefGoogle Scholar
  25. ———. 1960. What Is the Pure Theory of Law? Tulane Law Review 34: 269.Google Scholar
  26. ———. 1992 (1934). Introduction to the Problems of Legal Theory, Trans. B. Litchewski Paulson and S.L. Paulson. Oxford: Clarendon Press.Google Scholar
  27. Kramer, M. 2007. Objectivity and the Rule of Law. Cambridge: CUP.CrossRefGoogle Scholar
  28. Lamond, G. 2014a. The Rule of Recognition and the Foundations of a Legal System. In Reading HLA Hart’s ‘The Concept of Law’, ed. L. Duarte d’Almeida, J. Edwards, and A. Dolcetti. Oxford: Hart Publishing.Google Scholar
  29. ———. 2014b. Legal Sources, the Rule of Recognition, and Customary Law. American Journal of Jurisprudence: 1.Google Scholar
  30. MacCormick, N. 2008. Institutions of Law: An Essay in Legal Theory. Oxford: Oxford University Press.Google Scholar
  31. Moreso, J.J., and S. Chilovi. 2016. Il Diritto Come Linguaggio (nella Filosofia Analitica Contemporanea). In Che cosa e’il diritto. Ontologie e concezioni del giuridico, ed. G. Bongiovanni, G. Pino, and C. Roversi. Torino: Giappichelli.Google Scholar
  32. Munzer, S. 1972. Legal Validity. Leiden: Martinus Nijhoff.CrossRefGoogle Scholar
  33. Narváez Mora, M. 2015. Expressing Norms. On Norm-Formulations and Other Entities in Legal Theory. Revus. Journal for Constitutional Theory and Philosophy of Law 25: 43.Google Scholar
  34. Nino, C.S. 1978. Some Confusions Around Kelsen’s Concept of Validity. ARSP: Archiv für Rechts- und Sozialphilosophie 64: 357.Google Scholar
  35. ———. 1996. Introduzione All’analisi Del Diritto. Torino: Giappichelli.Google Scholar
  36. Peczenik, A. 1989. On Law and Reason. Berlin: Springer.CrossRefGoogle Scholar
  37. Perry, A. 2015. The Internal Aspect of Social Rules. Oxford Journal of Legal Studies 35: 283.CrossRefGoogle Scholar
  38. Pino, G. 1999. The Place of Legal Positivism in Contemporary Constitutional States. Law and Philosophy 18: 513.Google Scholar
  39. ———. 2014. Positivism, Legal Validity, and the Separation of Law and Morals. Ratio Juris 27: 190.CrossRefGoogle Scholar
  40. ———. 2016. Teoria analitica del diritto I: La norma giuridica. Pisa: ETS.Google Scholar
  41. Postema, G.J. 2011. Legal Philosophy in the Twentieth Century: The Common Law World. Vol. 11. Berlin: Springer.Google Scholar
  42. Prascina, E. 2009. Il Testamento Nuncupativo. Gazzetta Notarile: 527.Google Scholar
  43. Priel, D. 2011. The Place of Legitimacy in Legal Theory. McGill Law Journal/Revue de Droit de McGill 57: 2.Google Scholar
  44. Raz, J. 1979. The Authority of Law: Essays on Law and Morality. Oxford: Clarendon Press.CrossRefGoogle Scholar
  45. ———. 2011. From Normativity to Responsibility. Oxford: Oxford University Press.CrossRefGoogle Scholar
  46. Sandro, P. 2011. An Axiomatic Theory of Law. Res Publica 17: 343.CrossRefGoogle Scholar
  47. ———. 2019. Creation and Application of Law: A Neglected Distinction. Oxford: Hart/Bloomsbury, forthcoming.Google Scholar
  48. Sartor, G. 2000. Legal Validity as Doxastic Obligation: From Definition to Normativity. Law and Philosophy 19: 585.Google Scholar
  49. ———. 2008. Legal Validity: An Inferential Analysis. Ratio Juris 21: 212.CrossRefGoogle Scholar
  50. Tahko, T.E., and E.J. Lowe. 2015. Ontological Dependence. In The Stanford Encyclopedia of Philosophy, ed. E.N. Zalta. accessed 6 July 2016.
  51. Tuzet, G. 2014. Social Ontology and Legal Practice: A Debate. Rechtstheorie 45: 281.CrossRefGoogle Scholar
  52. von Wright, G.E. 1963. Norm and Action: A Logical Enquiry. London: Routledge.Google Scholar
  53. Walker, N. 2009. Out of Place and Out of Time: Law’s Fading Co-ordinates. Edinburgh Law Review 14: 13.CrossRefGoogle Scholar

Copyright information

© The Author(s) 2018

Open Access This chapter is distributed under the terms of the Creative Commons Attribution 4.0 International License (, which permits use, duplication, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, a link is provided to the Creative Commons license and any changes made are indicated.

The images or other third party material in this chapter are included in the work's Creative Commons license, unless indicated otherwise in the credit line; if such material is not included in the work's Creative Commons license and the respective action is not permitted by statutory regulation, users will need to obtain permission from the license holder to duplicate, adapt or reproduce the material.

Authors and Affiliations

  1. 1.Business SchoolUniversity of SalfordSalfordUK

Personalised recommendations