1In her recent article in Revus, Sara Azevedo deals with the topic of uninormative conflicts — “conflicts that stem from one (and only one) legal norm”.1 More precisely, the conflict in such cases arises between two particular norms that are both derivable from the same general norm.
- 2 Azevedo 2024: § 1 online.
2Azevedo provides two examples of uninormative conflicts. First, consider the case of Jodie and Mary:2
Take the case of Jodie and Mary, two conjoined twins whose separation, although necessary, would only guarantee the survival of one of the sisters. Although surgery would be the only way to ensure Mary's right to life, it would also interfere with Jodie's right to life.
- 3 Azevedo 2024: § 2 online
3Second, the vaccination case:3
Consider the case of COVID-19 vaccination campaigns. Initially, there were not enough doses to vaccinate the entire population, so it was up to the States to decide which groups to prioritise. The norm that enshrines the right to health gives rise to different prima facie duties falling on States — the duty to vaccinate A, B, C, D, and so on. Once again, normative authorities must determine which duty prevails at any given time, by issuing norms that define the priority groups for vaccination. However, vaccinating one group first and another last implies a limitation of the latter's fundamental right to health, even if constitutionally justified.
4This article focuses on two of Azevedo’s main claims. First, that uninormative conflicts are conflicts in abstracto (hereinafter: the classification claim). Second, a compound claim that uninormative conflicts need to be resolved, and that balancing is, at least usually, the only way to resolve them (hereinafter: the resolution claim). This article aims to shed critical light on both claims.
5In the second section, I present Azevedo’s classification claim in more detail and the argument against it, while in the third section I present the resolution claim and the argument against it.
6The classification claim is the result of the following argument:
1. All normative conflicts are (reducible to) conflicts in abstracto.
2. Uninormative conflicts are normative conflicts.
∴ Uninormative conflicts are conflicts in abstracto.
- 4 The distinction is widely associated with the work of Riccardo Guastini, but it can also be found i (...)
- 5 Guastini 2011: 106.
- 6 Guastini 2011: 107; Navarro & Rodríguez 2014: 183.
- 7 Guastini 2011: 107.
7The first premise of the argument amounts to a rejection of the widely held distinction between normative conflicts in concreto and normative conflicts in abstracto, so let’s briefly recall that distinction.4 Conflicts in abstracto refer to situations where the conflict of two norms occurs because the two norms, having a conceptually overlapping antecedent, relate their antecedent with mutually logically incompatible legal consequences. For example, a normative conflict in abstracto is said to exist between the two norms N1 “Vehicles are not allowed to enter the military base” and N2 “Military vehicles are allowed to enter the military base.” If we take p to stand for an object being a vehicle, pr to stand for an object being a military vehicle and q to stand to stand for entering in the military base, N1 can be represented as “p→O⌐q”, whereas N2 can be represented as “pr→Pq”. Given the conceptual overlap of their antecedent, and the logical incompatibility of their consequents, such normative conflicts are said to be identifiable in abstracto, that is in the law-interpretation phase, regardless of the specific cases that might be subsumable under the antecedent of each particular norm.5 Conflicts in concreto, on the other hand, refer to a conflict between two norms with mutually conceptually compatible antecedents but logically incompatible consequents.6 For example, a normative conflict in concreto is said to exist between the two norms N3 “Vehicles are not allowed to enter the military base” and N4 “Military machinery is allowed to enter the military base” in situations where a piece of military machinery is also a vehicle — say a military truck. If we take p to stand for an object being a vehicle, r to stand for an object being military machinery, and q to stand to stand for entering the military base, N3 can be represented as “p→O⌐q”, whereas N4 can be represented as “r→Pq”. Given the conceptual independence of their antecedents, such normative conflicts are said to be identifiable only in concreto, that is, only in the law-application phase, after the subsumption of the individual case under the antecedent of norms has been carried out.7
- 8 Although kernels of the argument are found in Gavazzi (1959: 66-69) and Alchourrón and Bulygin (197 (...)
- 9 And, using the famous Rossian typology of conflicts, a total-total conflict. Thus, as Yzaguirre not (...)
8This prima facie distinction has been the target of an influential argument now primarily associated with Giovanni Battista Ratti.8 The argument, briefly put, runs as follows: under the assumption that the law of strengthening the antecedent holds as a valid rule of reasoning, there is no difference between the two kinds of normative conflicts, as all conflicts in concreto can be represented as conflicts in abstracto. Take N3 “p→O⌐q” and N4 “r→Pq”. By strengthening the antecedent of N3 with the property r, we obtain N3’ “p∧r→O⌐q”, and by strengthening the antecedent of N4 with the property p, we obtain N4’ “p∧r→Pq”. Thus, we have “p∧r→O⌐q” and “p∧r→Pq” – a conflict in abstracto.9
- 10 Azevedo 2024: sec. 3.1, §§ 19–21 online. Several authors identify the two by using the term “confli (...)
9Azevedo follows several authors in identifying conflicts in concreto with conflicts of instantiation, conflicts that arise from multiple instantiations of a single norm.10 Thus, given that conflicts in concreto can be represented as conflicts in abstracto, what we call conflicts of instantiation are, according to Azevedo, also reducible to conflicts in abstracto. Furthermore, uninormative conflicts, given that they arise from the multiple instantiations of a single general norm and thus seem at first to be conflicts of instantiation (and conflicts in concreto), are also reducible to conflicts in abstracto.
- 11 The term “conflicts of instantiation” and the distinction are usually associated with the work of J (...)
10However, the identification of conflicts in concreto with conflicts of instantiation is a mistake. To see this, let’s see how several authors have explicitly distinguished the two.11 Conflicts of instantiation occur when an agent cannot simultaneously comply with the consequences derived from one or multiple general norms. Importantly, this impossibility arises from a factual impossibility of compliance. Consider a debtor who has taken loans from two different creditors but now lacks sufficient funds to repay both. In this case, the applicable general norm can be represented as:
N: ∀x,y: (Axy → O(Pyx))
11That is, for all individuals x and y, if x is in relation A with y, then it is obligatory for y to perform action P – such as repaying a loan – in relation to x. Let a and c stand for individual creditors, b to stand for the debtor. The two individual norms requiring the return of the loan to a particular creditor can be represented as:
N’: Aab → O(Pba)
N’’: Acb → O(Pbc)
12The impossibility of fulfilling both obligations to repay the loans is factual, not logical, as there is no logical incompatibility between the consequents of two norms.
- 12 Both examples and the representation are from Navarro & Rodríguez 2014: 182-184
13Or suppose we have a norm N: “Emergency service doctors should treat all heart attack patients immediately”. Further suppose that on a particularly unlucky night, there is a single emergency service doctor on shift. In a scenario where two people are having heart attacks in different parts of the city, the doctor cannot fulfill his obligation due to factual impossibility.12
14The distinction between conflicts in concreto on one side, and conflicts of instantiation on the other, is summarized by Navarro and Rodríguez as follows:
- 13 Navarro & Rodríguez 2014: 183. Emphasis mine.
Situations such as this may be called conflicts of instantiation, because they involve individual norms logically derivable from at least one general norm, or different instances of the application of the same general norm, that generate a conflict because it is empirically impossible to jointly satisfy them. The fundamental difference between conflicts of instantiation and normative contradictions stricto sensu, even in the case in which the latter arise via certain facts [i.e., conflicts in concreto], is that in the case of normative contradictions the impossibility of satisfaction is logical (e.g., it is logically impossible for the same subject at the same time jointly to satisfy the norms Op and O p), whereas in the case of conflicts of instantiation the impossibility is factual (e.g., b cannot pay now both creditors, not because such a thing is logically impossible, but merely because she has run out of money to do so).13
- 14 Martínez Zorrilla 2016: 741. Emphasis mine.
- 15 See Martínez Zorrilla 2016: 739-741. Martínez Zorrilla’s point is that the distinction between what (...)
15In a similar vein, Martínez Zorilla distinguishes “contextual antinomies” (which we referred to as “conflicts in concreto”) and conflicts of instantiation. Contextual antinomies are described as “logical inconsistencies produced whenever the same individual action is subsumable under two or more logically independent generic actions which are deontically modalized in an incompatible way”.14 The author, furthermore, highlights that contextual antinomies should not be regarded as in concreto, in the sense of only being detectable or arising in the law application phase, or due to empirical incompatibility.15
16On the other hand, regarding the conflicts of instantiation, Martínez Zorilla states that:
- 16 Martínez Zorrilla 2016: 741. Emphasis mine.
In this type of conflict, it is empirically impossible to carry out all the actions or behaviours deontically modalized for the generic case at stake, without logical inconsistency. This includes all situations in which, even without an inconsistency (either generic [i.e. in abstracto in our terminology] or contextual [i.e. in concreto]), the empirical circumstances prevent the possibility of complying with all the applicable norms (including the use of permissions).16
- 17 Although he discusses conflicts of instantiation under the heading of conflicts in concreto, this p (...)
17Now, the problem with identifying conflicts in concreto with conflicts of instantiation lies in the fact that the Ratti-style argument laid out above, targeting the distinction between conflicts in concreto and conflicts in abstracto, does not hold for conflicts of instantiation.17
18To see this, we return to the debtor-creditor situation where the debtor is faced with two norms that he cannot simultaneously satisfy. Again, let a and c stand for individual creditors, b to stand for the debtor. The two individual norms requiring the return the loans to particular creditors can be represented as:
N’: Aab → O(Pba)
N’’: Acb → O(Pbc)
19Strengthening the antecedent of N’ with Acb and strengthening the antecedent of N’’ with Aab produces the norms:
N’*: Aab∧Acb → O(Pba)∧O(Pbc)
N’’*: Acb∧Aab →O(Pbc)∧O(Pba)
20As we can see, there is no logical incompatibility between the consequences of the two norms. If logical incompatibility is to arise, one would have to strengthen their antecedents with “Pbc→ ⌐Pba” or “Pba→ ⌐ Pbc” — a claim that the repayment of one loan implies the non-repayment of the other. However, this is precisely the statement of empirical incompatibility that cannot be verified in abstracto and has nothing to do with the properties of the generic cases that form the antecedent of norms.
- 18 Azevedo 2024: sec. 3.4, § 34 online.
21Thus, Azevedo’s claim that conflicts of instantiation are the same as conflicts in concreto and that uninormative conflicts “involve normative inconsistency (not just factual impossibility)”18 seems mistaken. Uninormative conflicts, given that they arise from the multiple instantiations of a single general norm, are conflicts of instantiation and thus are also not conflicts in abstracto.
22To see this even more clearly, let’s return to the example of conjoined twins Mary and Jodie. Here, two particular norms are involved: one that determines that impairing Jodie's life is forbidden, and one that determines that impairing Mary's life is forbidden. And just like other examples of conflicts of instantiation, the impossibility of satisfying the two particular norms stems from the (empirical) fact that surgery on one sibling will result in the death of the other. A change in the empirical facts — say, a technological advancement that allows both siblings to be separated successfully — would mean that this impossibility, for this particular case, no longer holds.
23The same can be said for the vaccine case. The reason why administering the vaccine to one group impairs the right to health of the other is based on the empirical fact that there is no sufficient vaccine for everyone and that there are no other equally viable ways of securing the health of those who do not receive the vaccine.
- 19 Several authors adopt this idea tacitly or explicitly. See Alchourrón 1991: 413; Iturralde Sesma 19 (...)
- 20 Ratti 2015: 170-173.
- 21 I am using the term legal science to refer to everything traditionally done by legal science/legal (...)
- 22 Alchourrón and Bulygin 1971: 6
24Now, why is the discussion of whether uninormative conflicts are a matter of logical incompatibility or simply empirical impossibility even important? It is a long tradition in legal theory to adopt a position that can be dubbed normative conflict logicism. According to it, “a normative conflict” refers only to situations where incompatibility between norms is logical, i.e., expressible as sentences “Op∧O⌐p” or “Op∧⌐Op”.19 At least one motivation for normative conflict logicism proceeds as follows. It is widely held that one of the appropriate tasks of judges, properly falling under “application of law”, is resolving normative conflicts. Furthermore, resolving normative conflicts is also one of the legitimate tasks of legal science.20 When this view is combined with the thesis that legal science21 should be a value neutral, non-political endeavour,22 and that adjudication should not be law-creation but merely law-application, we arrive at the following conclusion: the only conflicts that legal science (to be considered a science) and courts (to respect the separation of powers) should resolve are those whose ascertainment is a value-neutral task. Therefore, the proper task of legal science and adjudication is to resolve only those conflicts arising from logical incompatibility, as these are the only conflicts that can be considered 'proper' normative conflicts.
25The summary of this view is nicely expressed by Carlos Alchourrón:
Everyone will admit that inconsistent normative systems should, for the sake of practical usefulness, be changed in favor of consistent ones. Moreover, I think that it will be accepted without objections that the identification of the conditions under which a normative system is consistent or inconsistent is a logical and not an ethical enterprise.23
- 24 Many authors are skeptical about this. Sardo claims that conflicts of instantiation are not normati (...)
26Commenting on the logicist view of normative conflicts or the metajuriprudential views on legal science that it assumes is far beyond the scope of this paper. However, it seems clear that conflicts of instantiation, whether they are actual normative conflicts or not,24 are the type of conflicts that will only sometimes merit the conclusion that they should lead to the modification of the applicable set of norms.
27This brings us to the resolution claim.
- 25 The consequences for the defeated norm will depend on the method or criteria used to resolve the co (...)
- 26 See Chiassoni 2024: 107.
- 27 I am using the term 'conflict resolution criteria' broadly to refer to balancing, as well as the le (...)
28To understand this important point about conflicts of instantiation, let us first make a more general point. The expression “resolution of a conflict” refers to the defeat of one of the prima facie applicable norms.25 Furthermore, the resolution of the conflict is usually guided by some conflict resolution criteria (often called methods, norms, or principles) that indicate which norm should legally be selected.26 The key point is that resolving the conflict in a manner contrary to what the conflict resolution criteria require is considered a legal mistake or a misapplication of the law.27
29Let’s now revisit the debtor-creditor case. Even though the debtor finds himself in a tragic situation where he will certainly violate at least one of his obligations, this does not seem like a good reason to change the general norm that has led to the conflict or to resolve it by defeating one (or both) of his particular obligations. Debtors should still return all their debts, and this particular debtor should still return both of his loans.
30Things are different in the case of the emergency service doctor. Here, too, we might think that there is nothing problematic with the general norm that doctors should treat all heart attack patients immediately, but rather, that the problem lies in the insufficient number of doctors available to respond. Navarro & Rodriguez summarize this point:
- 28 Navarro & Rodríguez 2014: 183. Emphasis mine.
[The general norm] itself should not be seen here as defective; after all, the best solution to the problem seems to be to assign more doctors to the hospital. However, if such an alternative is in fact unattainable and this limitation is seen as axiologically relevant, then it would appear desirable to modify the normative system.28
- 29 Similar point is made by Spaak 2023: 10.
31The legislator might make this modification at the level of the legal system by changing the set of existing legal norms, thus preventing the conflict. Alternatively, it might be done at the level of law application by resolving the conflict. Be that as it may, the main point regarding conflicts of instantiation is that the decision to prevent or resolve them is grounded in the judgment that factual impossibility is axiologically inappropriate.29
- 30 Azevedo 2024: sec. 4.1, § 42 online.
32With all this in mind, we return to the resolution claim that uninormative conflicts need to be resolved and that balancing is the only way to solve them. The first part of the claim — that uninormative conflicts need to be resolved — is based on two premises. First, there is a seemingly conceptual premise that, in the case of uninormative conflicts, legal operators face "a set of alternatives — e.g., applying norm1 or norm2."30 Second, there is a claim that, given the prohibition of non liquet, legal operators must choose one of the norms that governs the case.
- 31 The point here is mostly one of clarification rather than criticism. I do not claim that making nor (...)
33In light of our previous discussion, the status of the first premise is clear. In uninormative conflicts, the assertion that the two individual norms are alternatives — in the sense that they are mutually exclusive — is a disguised normative claim; specifically, it suggests that the factual impossibility that occurs in the case is normatively (morally, politically) unjustified and that only one of the norms should be selected as applicable, all things considered. In other words, it should be considered a legal mistake to hold that both norms are applicable.31 Furthermore, the prohibition of non liquet does not require judges to select just one of the norms; rather, it excludes the possibility of a judge refusing to make a choice. If the judgment of axiological impropriety is not made, the judge faces a choice among three options: applying the first individual norm, applying the other individual norm, or applying both.
- 32 The claim is a specific version of the residual nature of the applicability of balancing, whereby b (...)
34The second part of the resolution claim — that balancing is the only way to resolve uninormative conflicts — is again based on two premises. First, a claim that the traditional criteria for resolving conflicts (lex posterior, lex specialis, and lex superior) are conceptually inapplicable to uninormative conflicts. Second, that as a matter of positive law in most legal systems, no other conflict resolution criterion is available to be used in cases of uninormative conflicts.32
- 33 Azevedo 2024: sec. 4.1, § 44 online
- 34 The point, therefore, is that uninormative conflicts are not really uninormative.
35It should be highlighted that both of these claims are correct. However, we should focus on Azevedo’s explanation of why the first claim regarding the conceptual inapplicability of the three criteria holds. The contention is that the three traditional criteria are inapplicable because they “they rely on the properties that one norm exhibits in relation to another”,33 while in uninormative conflicts the conflict is, supposedly, generated only by one norm.34
- 35 Here, I refer to the traditional understanding of lex superior as a formal criterion, employing the (...)
36But this claim is a misdiagnosis as it ignores the author’s own important insight: that even in uninormative conflicts we have a conflict between two particular norms. Thus, the reason why lex specialis and lex superior are not applicable is not that there are not two norms, but rather that the two norms in question are hierarchically equal and neither of them is more special than the other.35
- 36 The point that follows holds even if one takes this individual norm to arise at a point that preced (...)
- 37 This follows if one adopts a very minimal, i.e., “thin”, definition of the lex posterior principle (...)
- 38 Alchourrón & Bulygin 2015: 326.; Agüero San Juan 2017: 234
- 39 The minimal account of the lex posterior mentioned above seems to fail to capture this important co (...)
37Things are more interesting with the lex posterior criterion. The duty of the state (an individual norm) to protect the life and health of a person (usually) arises only after the person is born.36 Thus, two particular norms obliging the state to protect can come about at different times. It seems, prima facie, that nothing prevents the application of lex posterior criterion.37 However, even this criterion should be considered conceptually inapplicable. The resolution of normative conflicts by the use of lex posterior criterion is conceptually linked with the capacity of the normative authorities to enact deliberate normative change, i.e., change the normative status of certain actions.38 If this is so, then the only norms to which the lex posterior criterion conceptually apply are those that differ with regard to the time from which they belong to a normative set. This is because they are the results of different acts of deliberate normative change, or derived from such results. Even if the individual norms stemming from one general norm arise at different times, they do not do so because they are enacted by the normative authority at different times. Instead, they are instantiations of the same generic case that have emerged at different times. Lex posterior is thus not applicable to uninormative conflicts.39
38To conclude, let’s take stock of the results. The classification claim, which posits that uninormative conflicts are conflicts in abstracto, has been challenged by demonstrating that these conflicts arise from factual impossibilities rather than logical inconsistencies. By clarifying the difference between conflicts in concreto and conflicts of instantiation, I argued that uninormative conflicts, rooted in the multiple instantiations of a single general norm, are not a matter of logical incompatibility.
39This result had a direct bearing on the resolution claim — that uninormative conflicts must be resolved through balancing. It was argued that this claim is a normative claim, justified by the tacit assumption of axiological unacceptability of a certain case uninormative conflict. Finally, while Azevedo is correct about her basic point regarding the inapplicability of traditional conflict-resolution criteria, my argument offers a different rationale for this inapplicability. It is grounded in the fact that the two norms in question are equivalent in terms of their hierarchy and specificity, as well as in the conceptual link between lex posterior and the enactment of normative change.