• 1 Austin [1961] 1970: 182 (emphasis in original).

When we examine what we should say when, what words we should use in what situations, we are looking again not merely at words (or “meanings,” whatever they may be) but also at the realities we use the words to talk about: we are using a sharpened awareness of words to sharpen our perception of, though not as the final arbiter of, the phenomena.1

• 2 Bobbio 1965: 321.

1The thesis according to which coercion, or recourse to force, is a necessary feature of law is characteristic of a traditional definition of law as “an organized body of coercive rules.”2

• 3 Bobbio 1965: 321.

2There are, in fact, two possible versions of this thesis, as Norberto Bobbio has clarified. In the traditional view, coercion, or recourse to force, is a necessary feature of law because it is considered “as a necessary means for [its] realization”; according to a second, more recent, view, coercion is instead considered as the distinctive content of legal rules.3

• 4 Our translation and emphasis. The German original: “Recht ist der Inbegriff der mittelst äußeren Zw (...)

3The traditional view is exemplified by Rudolf von Jhering’s definition of law: “Law is the complex of rules of conduct, maintained by a political authority by means of external coercion in order to secure the essential conditions of life.”4 According to the traditional view, legal norms can be distinguished from other social norms in virtue of the fact that they are backed by the threat of coercion in the event of non-compliance.

• 5 Bobbio 1965: 321–322. A similar view has been maintained, according to Bobbio, by Karl Olivecrona a (...)
• 6 Bobbio 1965: 334 (emphasis added).

4To the traditional view, Bobbio contrasts the more recent view, originally proposed by Hans Kelsen, according to which “law is not a body of rules guaranteed by force, but a body of rules about force.”5 What distinguishes a legal system from other social normative systems, such as religious and moral orders, is “the presence not of sanctions, and therefore of rules that are sanctioned, but of rules that regulate sanctions.”6

• 7 Kelsen 1945: 116 (emphasis added).

5This idea is famously expressed by Kelsen in the formulation of the basic norm itself (the Grundnorm) of a legal system: “Coercive acts ought to be carried out only under the conditions and in the way determined by the ‘fathers’ of the constitution or the organs delegated by them.”7

• 8 Kelsen 1945: 19.

6For Kelsen, the law is a specific social technique: the social technique of a coercive order, which “consists in bringing about the desired conduct of men through the threat of a measure of coercion which is to be applied in case of contrary conduct.”8

• 9 Kelsen 1945: 21.

7Kelsen underlines that the law paradoxically threatens the use of force in order to forbid the use of force among the members of the community. However, legal norms attach specific conditions to the use of force in relations among men, thus authorizing the employment of force only by certain individuals and only under certain circumstances. Law thus “makes the use of force a monopoly of the community ..., and precisely, by so doing, law pacifies the community.”9 It is in this specific sense that, according to Kelsen, coercion is the content of legal norms and law is a coercive order.

• 10 Bobbio 1965: 335.

8However, Bobbio remarks that the theory according to which coercion is the content of legal norms concerns “not the single rules, but the system in its entirety, and consequently the definition of law amounts not to a criterion for distinguishing a legal rule from one that is not legal, but a legal system from other non-legal systems.”10

• 11 Himma 2020: 1ff.

10Similarly to the traditional theory, Himma understands coercion as a means for the realization of law: coercion, for Himma, is not directly the content of mandatory legal norms governing non-official behaviour; it is threatened as a means to deter and punish non-compliance with those norms.11

• 12 As Bobbio (1965: 324ff.) remarks, one of the main objections raised against the traditional theory (...)

11However, similarly to Kelsen, Himma maintains that coercion is a necessary feature not of single legal norms, but of a system of law in its entirety. According to Himma, for a normative system to count as a system of law it is not necessary that all the system’s norms are backed by a coercive sanction, but rather that at least some of its norms (and, more specifically, some of “the mandatory norms governing non-official behaviour”) are backed by a coercive sanction.12

• 13 Himma 2020: vi and n. 2. Himma (2020) has Kelsen (1945) say that the only criterion for distinguish (...)

13What distinguishes legal systems from other social normative systems, such as moral and religious orders, is not, for Kelsen, coercion as such; it is rather “socially organized sanction.”13

• 14 Kelsen 1945: 16.

14Kelsen recognizes that “every social order is somehow ‘sanctioned’ by the specific reaction of the community to conduct of its members,”14 and this is also true of moral systems and religious systems. However, legal systems differ from other social orders in virtue of two peculiar features of legal sanctions:

• 15 Kelsen 1945: 16.

16(ii) in contradistinction to moral sanctions, which “consist in the automatic reaction of the community not expressly provided by the order,” sanctions provided by legal systems are socially organized, that is, they are expressly provided by the legal order as definite sanctions that are to be applied only under the circumstances and by the organs determined by the legal order.15

• 16 Kelsen 1945: 20.
• 17 Bobbio 1965: 330–331.

17Kelsen expressly remarks also that “other social orders pursue in part the same purposes as the law, but by quite different means”: both law and morality, for instance, forbid murder; but “the law does this by providing that if a man commits murder, then another man, designated by the legal order, shall apply against the murderer a certain measure of coercion, prescribed by the legal order.”16 On the contrary, “morality limits itself to requiring: Thou shalt not kill.” Kelsen recognizes that a murderer may also be “morally ostracized” by his fellow men; but, according to Kelsen, “the moral reaction against immoral conduct is neither provided by the moral order, nor, if provided, socially organized,” whereas the reaction of the law consists in “a measure of coercion enacted by the order, and socially organized.” Bobbio recapitulates Kelsen’s concept of law as follows: “Law is the ensemble of the rules or norms that regulate the when, the who, the how, and the how much in the exercise of coercive power.”17

• 18 Kelsen 1942: 16.
• 19 Kelsen 1945: 19.

20When Kelsen asks: “What is the essence of law?”, he is not analysing a concept; he is rather elaborating a concept that can give an answer to questions such as “What is the criterion by which law can be distinguished from other social forms?”18 and “What could the so-called law of ancient Babylonians have in common with the law that prevails today in the United States?”19

• 20 Kelsen is not unaware of Kant’s “Copernican revolution”: the clarification of the transcendental ro (...)

21Despite the fact that Kelsen starts from the common usage of the word “law,” his inquiry is not an investigation of the mere semantics of the word “law” as it is used in a specific language (English, for instance) in a specific community; it is an investigation of the phenomena that are designated by the word “law.” His inquiry is directed not so much to the meaning of the word “law,” as to its denotation; not so much to the concept, as to the structure of phenomena for which a proper concept is to be found.20

• 21 We use here the adjective “phenomenological” to generally refer to an investigation oriented to phe (...)

22The “touchstone” for evaluating Kelsen’s theory is thus not “ordinary talk” about legal phenomena; it is rather legal phenomena themselves. Kelsen is not doing a conceptual analysis, but rather a phenomenological analysis: an analysis that seeks to elaborate a concept that fits and accounts for existing phenomena.21

• 22 Himma distinguishes modest conceptual analysis from what he calls immodest conceptual analysis draw (...)

24The novelty of Himma’s philosophical enterprise consists in the specific methodology he adopts. Himma’s intent is to defend the Coercion Thesis exclusively on the basis of conceptual analysis, and more specifically on what he calls modest conceptual analysis, which is articulated in two steps.22

• 23 Himma 2020: 46. Surprisingly, though, the English dictionaries Himma relies upon are not the unabri (...)

25(3.1) The first step in modest conceptual analysis is what Himma calls an “empirical observation” of “ordinary intuitions” as they are manifest in the “contingent linguistic conventions for using the relevant concept-term” in “ordinary talk” (with the exclusion of the possibly more rigorous linguistic conventions of the academic and philosophical community). For this first step – which seems to be sociological in character – Himma relies upon the definitions that can be found in English dictionaries, which, according to Himma’s oxymoron, “roughly but accurately” express the lexical meaning of the relevant concept-terms “as it is determined by the canons of ordinary usage governing its use.”23

• 24 Lakoff [1987] 1990: 75–77.

26Starting from dictionary definitions is undoubtedly a useful and fruitful practice in philosophical research. However, a philosophical inquiry that starts from a survey of dictionary definitions should not forget that, as George Lakoff remarks, “the human beings who write dictionaries vary in their choices” and “[t]hough choices made by dictionary-makers are of no scientific importance, they do reflect the fact that, even among people who construct definitions for a living, there is no single, generally accepted cognitive model, even for such a common concept as ‘mother.’”24

• 25 Austin [1961] 1970: 183. It is well known that in ordinary language practices many proverbial expre (...)

27As J. L. Austin remarks, one of the possible snags in the analysis of ordinary language is the “snag of Loose (or Divergent or Alternative) Usage.” Austin asks himself: “Do we all say the same, and only the same, things in the same situations? Don’t usages differ?”25

• 26 Himma 2020: 54.

28Himma is aware of this snag, but he remarks that all the different possible conceptual models associated with the different ordinary meanings of the word “law,” as referred to normative systems, support the Coercion Thesis. Nonetheless, he chooses to narrow his investigation down to systems of municipal law, since they “constitute the paradigm cases conditioning the canons of ordinary usage with respect to the concept-term law reported in dictionary definitions.”26

• 27 Himma 2020: 46.
• 28 Himma 2020: 47.

30(3.2) The second step in modest conceptual analysis consists in identifying “the philosophical assumptions about the metaphysical nature of a thing to which the corresponding concept-term refers.”27 This second step is not merely empirical and descriptive; it rather involves a logical analysis that Himma likens to “the project of proving some mathematical theorem in a manner satisfying the favored logical axioms.”28 The identification of shared philosophical assumptions, then, is not merely a survey; it is intended to explicate the relevant assumptions in a more determinate and consistent way than ordinary talk and linguistic conventions usually exhibit, and to make explicit their implications, in order to overcome possible conceptual problems.

• 29 Himma 2020: 43.
• 30 Himma 2020: 43 (emphasis added).

31It is in the explication of these philosophical assumptions, and in the derivation of their logical and metaphysical implications, that Himma fully displays the keenness of his conceptual analysis and of his argumentations. According to Himma, “[c]onceptual problems arise only insofar as [linguistic] conventions are either indeterminate or facially inconsistent with respect to some entity of philosophical interest.”29 Modest conceptual analysis is purported to “explicate the underlying philosophical assumptions grounding these linguistic conventions with sufficient depth and detail to resolve the indeterminacy or inconsistency motivating the inquiry.”30

• 32 Bobbio [1950] 1997: 32.

33Linguistic analysis, though, is expressly conceived by Bobbio not as merely descriptive or explicative but as critical: “Ordinary language is rendered more rigorous and less flexible, or indeed entirely supplanted by, scientific language.”32

• 33 Himma 2020: 39. Himma does not univocally determine whose philosophical assumptions must be taken i (...)

35It is important to remark that these philosophical assumptions are not those of academic philosophers, but rather what Himma calls “our philosophical assumptions,” where it is unclear whom the adjective “ours” refers to – except for the exclusion of the academic philosophical community, whose possible “best theories” are not relevant for Himma, because “what matters is what people believe.”33

• 34 Austin [1961] 1970: 183.

36This approach seems to expose Himma to a second typical snag highlighted by Austin in the analysis of ordinary language: the “crux of the Last Word.” The problem raised by Austin is this: “Why should what we all ordinarily say be the only or the best or final way of putting it? Why should it even be true?”34

• 36 To be true, Himma seems to violate his own epistemic standards when he admits that his explication (...)
• 37 Himma 2020: 37.

38All a philosopher can do with regard to a concept is describe how that concept is contingently used in ordinary talk and explicate people’s philosophical assumptions that are contingently related to that concept.36 Therefore, modest conceptual analysis is, for Himma, “the only epistemically viable approach to conceptual analysis.”37

• 38 See Himma 2020: 39–40.

40Modest conceptual analysis, in fact, rather than an analysis of concepts, appears to be an analysis of historically and culturally contingent conceptions, like the analysis that can be made of the historically and culturally contingent conception of combustion positing the necessary presence of phlogiston.38 In setting out to make a modest conceptual analysis of the conception of combustion based on phlogiston, it would be inconsistent with the aforementioned methodological presuppositions, and even absurd, to not restrict oneself to describing that conception, but to also defend it.

• 39 See Himma 2020: 1.

41Nonetheless, Himma expressly asserts that his book defends the Coercion Thesis.39 This appears to be a methodological leap, and certainly again raises the crux of the Last Word, which Himma attributes to ordinary talk, the common usage of words, and shared philosophical assumptions, while he expressly rules out the relevance of more rigorous or refined analysis of empirical phenomena made in the context of academic research.

• 40 It is well-known, for instance, that there are many languages in which the term “law” does not have (...)

43The first question is: Can the results of such a conceptual analysis also be valid for non-American English-speaking people, or for non-English speaking people?40

• 42 Himma 2020: 8.

48Both the forward- and backward-looking aspects of punishment are linked to the fact that punishment is necessarily “norm-governed.” The backward-looking aspect consists in the retributive element, which presupposes that punishment “is justified under some set of norms” belonging to the same system of norms “dictating what the subject must do to avoid punishment.”42 But insofar as punishment is norm-governed, it also has, for Himma, the forward-looking aspect of deterrence against non-compliance based on the threat of a detriment in case of non-compliance with a specific substantive norm.

• 45 Himma 2020: 8.

52These two claims are mitigated by Himma’s remark that, in some contexts, “depending on the content of the relevant moral or legal norms, something done as an act of retaliation ... might be morally or legally justified.”45

• 46 The first definition (not considering two preceding definitions concerning an obsolete meaning) is (...)

54From a lexical point of view, by consulting the unabridged online edition of the Oxford English Dictionary, one would see not only that the term “retribution” appears in the definiens of “retaliation” in the first definition of the word,46 but also that a third definition expressly refers to the “law of retaliation,” that is, “a retributive form of justice whereby an offender’s punishment resembles the offence committed in kind and degree.”

• 47 See Kelsen 1942 and 1943 (remarking, for instance, that only the existence of an obligation to reve (...)

57Many historical and anthropological studies of vindicatory systems have shown, indeed, that retaliation is not so much motivated by a desire for revenge as it is normatively imposed as a social and legal duty. Kelsen, for instance, considers blood revenge the first form of socially organized sanction and underlines the social preventive effect of revenge: he accordingly identifies vindicatory systems as the first systems of law.47

• 48 According to Terradas Saborit – and contrary to a widespread view – revenge in vindicatory systems (...)

58Furthermore, recent anthropological research has drawn attention to the necessity to distinguish a vindicatory paradigm of justice from the penal paradigm of justice and highlights that, contrarily to a widespread view, vindicatory systems of law are “primarily based on the compositional processes” – that is, on legal processes aimed at coming to an agreement on compensation for an injury or insult. Revenge, in such systems, only has a “subsidiary validity” and “actually appears, both ethnographically and historically, ruled by a judicial authority” (see Terradas Saborit forthcoming 2021).48

• 49 One could also remark that, contrary to what Himma maintains, an act of retaliation may “communicat (...)

59The semantic and anthropological recognition of the actual features of vindicatory systems suggests that, from a conceptual point of view, retaliation can be considered an intrinsically norm-governed act: retaliation is, indeed, both a normatively determined and a normatively authorized, or even a normatively imposed, sanction.49