The Limits of Deduction in the Identification of Customary International Law

Abstract

Much scholarship on customary international law has examined the merits of induction, deduction, and assertion as approaches to custom identification. Save for where international tribunals identify custom by assertion, writers have viewed custom identification that does not rely on evidence of State practice and opinio juris as an example of deductive reasoning. However, writers have stated that, at best, deduction is reasoning from the general to the particular. This article draws on legal philosophy to define the contours of deductive reasoning and argues that pure deduction, namely deduction not combined with other forms of reasoning, is an unsound approach to custom identification. This argument is tested by reference to cases of custom identification by the International Court of Justice, categorised according to three types of deduction: normative, functional, and analogical. This article also explores the authority and utility of custom identification by pure deduction and its impact on content determination.

One can rely on different forms of reasoning to ascertain the existence of rules of customary international law in a process called “identification”.¹ In its 1969 judgment in North Sea Continental Shelf, the International Court of Justice (ICJ or the Court) stated that, for customary rules to be formed, State practice should be “extensive and virtually uniform” and should occur “in such a way as to show a general recognition that a rule of law or legal obligation is involved”.² North Sea Continental Shelf exemplifies custom identification by inductive reasoning, where one assesses whether there is sufficient evidence of the two elements of custom, State practice and opinio juris, to find whether new customary rules exist.³ In its 1986 judgment in Nicaragua, the Court held that the rule on non-intervention was “a corollary of the principle of the sovereign equality of States”.⁴ Nicaragua exemplifies custom identification by deductive reasoning, where one does not consider evidence of State practice and opinio juris. Still, it infers the existence of new customary rules from other established rules. This article discusses the limits of deduction as a form of reasoning to identify rules of customary international law. Following the terminology of the International Law Commission (ILC), this article calls the various forms of reasoning “approaches”.⁵

Nicaragua sparked an intense scholarly debate on how the ICJ identifies customary international law.⁶ In the eyes of many, this debate was settled in 2015 when Talmon wrote an article arguing that the induction/deduction dichotomy paints a distorted or incomplete picture of the ICJ's approach to custom identification because the Court generally only states what customary international law is, without inductive or deductive reasoning to support its statements.⁷ This approach has become known as “assertion”. In his article, Talmon also discussed deductive reasoning, writing that the Court uses three types of deduction: normative, to infer new rules from established ones;⁸ functional, to infer new rules from the functions of persons or organizations;⁹ and analogical, to extend the rationale of existing rules to cases not covered by them by reason of a “common cause or link between the two situations”.¹⁰ Deduction can help “confirm and strengthen” the results reached by induction or “make up for a less than comprehensive or conclusive inductive process”.¹¹ The scholarly debate on approaches to custom identification overlapped with the ILC's work on Identification of Customary International Law, which started in 2012 and was completed in 2018. The 2018 Draft Conclusions, of which the General Assembly has taken note,¹² set out the principles governing identification as the process to ascertain the existence of customary international law.¹³ The ILC referred to deduction only in its commentary to Draft Conclusions 1 and 2, which stated that identification based on evidence of State practice and opinio juris does not identify customary rules by “‘deductive’ approaches”. However, the former does not preclude some “deduction as an aid”. ¹⁴

Scholars accept that pure deduction, that is, deduction not combined with other forms of reasoning, can be a sound approach to identify customary international law. This view builds on the assumption that one knows what reasoning by deduction is. Most works do not explain what their authors understand deductive reasoning to be, beyond stating that it is reasoning from the general to the particular.¹⁵ Talmon seems alone in acknowledging the link between deductive legal reasoning and classical deductive logic. Still, he stated that the former has evolved independently of the latter because of its connection with what he calls, without elaboration, “the logic of the law”.¹⁶ Owing to this assumption, scholars have accepted that any approach to custom identification, except assertion, that does not exhibit the hallmarks of induction, namely more or less detailed references to evidence of State practice and opinio juris, is an example of custom identification by deductive reasoning.

Building on a clarification of deductive reasoning, this article argues that pure deduction is inadequate to identify customary international law. It does not argue that deduction is never helpful to identify customary rules but that pure deduction cannot support identifying such rules. Some induction is required to ground custom identification in “good justifying reasons”.¹⁷ By exposing the problems of pure deduction, this article is an indirect plea for identifying customary international law by induction or, at least, a combination of induction and other forms of legal reasoning. In developing its argument, this article recognizes that the ICJ may use distinct forms of reasoning in one custom identification exercise. The Court may state that evidence of State practice and opinio juris confirms the existence of a customary rule but may support its statement by deductive reasoning instead of listing that evidence.¹⁸ In this article, cases in which the Court adopts this modus operandi are seen to be custom identification by induction for two reasons: first, the Court avows that it identifies customary rules based on evidence of State practice and opinio juris, although it does not give much detail; second, other research shows that the constraints of the ICJ's deliberation process result in its judicial decisions recording custom identification carried out by induction as reasoning which does not detail evidence of State practice and opinio juris.¹⁹

This article focuses on the jurisprudence of the ICJ becasue it is the only international court of general jurisdiction which, unlike other international courts and tribunals created by specialized treaty regimes, commonly identifies rules of customary international law to decide cases before it. The article is divided into four sections. Section I formulates an underlying framework: it explains what legal reasoning is, defines the features of deductive legal reasoning and explains what forms of legal reasoning can persuasively support custom identification. In Section II, this article tests whether purely deductive legal reasoning can persuasively justify the finding that customary rules exist. The test cases are decisions in which the ICJ has identified customary rules by pure deduction, and the framework is Talmon's distinction between normative, functional and analogical deduction. Section III discusses wider implications concerning the authority of custom identification by pure deduction, its utility, and its impact on content determination. Section IV concludes.

I. Legal Reasoning, Deduction and Custom Identification

To assess whether purely deductive reasoning can support custom identification, one should proceed in stages: first, define what legal reasoning is; second, elucidate the characteristics of deductive legal reasoning; and third, explain which forms of legal reasoning can persuasively ground the identification of custom conceived as the union of State practice and opinio juris.

A. Legal Reasoning as Persuasion

Legal reasoning is a process of argumentation with the practical aim of persuasion.²⁰ Underlying this practical aim is a justificatory function: the aim of legal reasoning is to provide a public basis to justify arguments before courts of law and the decisions of those courts.²¹ Judicial reasoning, such as reasoning to support findings that customary rules exist, must “actually support the decision to the extent that it shows why the order given […] is justified”.²² Courts “actually support” their decisions by grounding them in “good justifying reasons”.²³ What constitutes “good justifying reasons” is assessed by the court's audience. For example, depending on whether the ICJ supports decisions that customary rules exist by grounding them in good justifying reasons, it may succeed or fail in persuading that those customary rules exist. If the Court fails, one should question whether the customary rules not identified by reference to good justifying reasons rest on sufficiently solid foundations. Whether the Court has succeeded in providing good justifying reasons is assessed by the audience of the Court's decisions, constituted primarily of States, and can include international organizations, the community of scholars, and individuals affected by those decisions.²⁴

Within the framework of legal reasoning thus understood, one can use different forms of reasoning to achieve the objective of persuading by way of good justifying reasons. There are three principal forms of reasoning commonly used in legal propositions: inductive, deductive and analogical. Each form relies on certain factors peculiar to it to provide good justifying reasons. The premises of inductive reasoning do not compel any specific conclusion, but inductive reasoning derives its strength from two aspects: first, from its foothold in empirical observation;²⁵ second, where different empirical premises justify identical propositions, from recognizing that there is a more remote and general premise justifying such propositions.²⁶ One could observe that on a particular day, the sky is covered in dark clouds and that it is windy, from which one could infer that it will rain. First, empirical observation of dark clouds and wind is the premise of an inductive proposition and suggests the conclusion that it will rain. Second, this conclusion builds on our knowledge of other instances in which certain sky and wind conditions have foreshadowed rain, from which one can recognise a more remote and general premise that rain will likely come when the sky is covered in dark clouds and it is windy. This conclusion is not compelled by its premises but derives its persuasiveness from observation of real-world phenomena and abstraction of more general premises from them.

Deductive reasoning derives its persuasiveness from the possibility that certain premises compel a certain conclusion.²⁷ The classic example of deductive reasoning concerns the mortality of Socrates. Given the two premises that all men are mortal and Socrates is a man, it must follow that Socrates is mortal. Deductive propositions are constructed as syllogisms: each syllogism has two premises from which a conclusion can necessarily follow. However, constructing propositions as syllogisms does not guarantee that their premises will compel a certain conclusion.²⁸ One can divide deductive propositions into two categories: propositions the premises of which compel a certain conclusion, called “perfect syllogisms”; and propositions the premises of which do not compel a certain conclusion, called “imperfect syllogisms”.²⁹ These types of syllogisms are further discussed below.

Similar to inductive propositions, the premises of analogical reasoning do not compel a certain conclusion.³⁰ Analogical reasoning builds on the existence of two situations: a prior one, which is already governed by certain rules or principles, and a posterior one, yet unregulated. Reasoning by analogy compares these two situations to justify extending the rules or principles governing the prior situation to the posterior one. The strength of reasoning by analogy rests on two factors: how close the analogy between prior and posterior situations is and how authoritative the rules or principles governing the prior situation are.³¹ One may know that if one puts a pot of water over a source of heat, the water will eventually boil. One may also conclude that a pot of oil will boil if put over a source of heat because of the similarity of this situation with the one involving a pot of water. The persuasiveness of this reasoning by analogy rests on the fact that water boils when heated up and on the similarity between the two situations, both of which involve a pot of liquid put over a source of heat. Analogical reasoning is typical of common law systems, where judges apply precedents on the strength of analogies between regulated and unregulated cases.³² This form of reasoning is known to international law, where some have used it, for example, to determine the applicability of rules of international law to cyberspace.³³ Owing to its reliance on earlier authorities, reasoning by analogy builds on the historical evolution of rules, principles and precedents.³⁴ One can thus compare it to inductive reasoning.³⁵

Inductive, deductive and analogical reasoning need not be applied separately but can be combined to provide good justifying reasons to support the conclusions of judicial decisions.³⁶ The example of the boiling pot of oil is framed as an analogy, but it also rests on the empirical observation that pots of water boil when put over sources of heat. Judicial decisions may not always be as persuasive as they could be. There should be little doubt that judges seek to use the most persuasive justifications available to them to ground their decisions, whether by using forms of reasoning in isolation or by combining them. Either way, legal reasoning at least aspires to give entirely persuasive justifications.³⁷

B. Characters of Deductive Legal Reasoning

The question arises as to whether deduction is a form of reasoning that can provide judicial decisions with persuasive justifications. The answer lies in the distinction between perfect and imperfect syllogisms and in the formal validity and truth of legal syllogisms.

1. Perfect and imperfect syllogisms

In logic, deductive propositions take the form of syllogisms built on a major and a minor premise. Syllogisms are perfect if conclusions follow from the stated premises alone and imperfect if they need other information additional to that in the major and minor premise to compel their conclusions.³⁸ Given a norm under which “everyone must shake hands when meeting for the first time” and a case where “Harry meets Sally for the first time”, it follows that “Harry and Sally must shake hands”. In this perfect syllogism, the norm is the major premise and the factual situation to which that norm applies is the minor premise.³⁹ The combination of a major and minor premise compels the conclusion without the need for additional information. This syllogism would be imperfect if its major premise were that “everyone must shake hands when meeting in the evening” and its minor premise that “Harry meets Sally at 5:45 p.m.”. The conclusion that “Harry and Sally must shake hands” only follows necessarily from these premises if, in the location where Harry meets Sally, 5:45 p.m. is recognized as part of the “evening”.

Syllogisms are perfect only if their premises cannot be demonstrated, meaning they do not have to be supported by reasoning as their audience accepts them without further explanation.⁴⁰ The mortality of humans can be demonstrated, for example, by empirical means. Still, Socrates's syllogism is perfect because its audience accepts that humans are mortal without the need for further explanation.⁴¹ Syllogisms can be perfect, and deductive reasoning can thus be entirely persuasive, even if their premises can be demonstrated.

Legal reasoning aspires to provide entirely persuasive justifications as to whether legal systems attach certain consequences to the existence of certain facts. Perfect syllogisms fulfil this aspiration because their premises compel certain conclusions. Conversely, imperfect syllogisms fulfil this aspiration to a much lesser extent, as their premises cannot compel their conclusions. From a rule under which “a State's territorial sea can extend up to 12 nautical miles (nm) from that State's coast” and a case where “Italy is a coastal State”, it necessarily follows that “Italy's territorial sea can extend up to 12 nm from its coast”. The aspiration of legal reasoning to provide entirely persuasive justification is inconsistent with the imperfect syllogism built by the same minor premise, that “Italy is a coastal State”, and a different major premise, that “a State's territorial sea can extend up to 12 nm”. The conclusion that “Italy's territorial sea can extend up to 12 nm” is silent as to the point from where one measures the 12 nm.

2. Formal validity and truth of legal syllogisms

Whether legal syllogisms can be entirely persuasive depends on their formal validity and truth. Syllogisms are formally valid if the premises compel the conclusions, but their formal validity does not depend on the truth of the premises. A conclusion follows from certain premises irrespective of their truth, which is a separate, empirical question.⁴²

The formal validity of a syllogism also does not imply the truth of its conclusion. If the major premise were that “a State's territorial sea can extend up to 6 nm from its coast” and the minor premise were that “Italy is a coastal State”, the conclusion that “Italy's territorial sea can extend up to 6 nm from its coast” would be formally valid as a matter of logic. The fact that, in international law, territorial seas could extend up to 12 nm from a State's coast would not affect the formal validity of this syllogism. Instead, that fact would make the major premise false, thus affecting the truth of the conclusion compelled by its combination with the minor premise. The same would take place if the minor premise were false. If the major premise were that “a State's territorial sea can extend up to 12 nm from its coast” and the minor premise were that “Mongolia is a coastal State”, the conclusion that “Mongolia's territorial sea can extend up to 12 nm from its coast” would be formally valid as a matter of logic. The formal validity of this syllogism would be unaffected by the fact that Mongolia is landlocked, while the false minor premise would affect the conclusion's truth. One verifies the truth of the major and minor premise through empirical inquiry into, respectively, how wide the territorial sea is under international law and whether the relevant State has a coast.

The truth, or conclusiveness, of legal syllogisms does not depend only on the truth of their premises, but also on the validity of the rules used in their premises. This “validity thesis” recognizes that a legal system has criteria to determine which rules within the system are “presumptively sufficient” to be “valid rules”. In English law, one can summarise these criteria by stating that one recognizes as law “whatever the [King] in Parliament enacts”.⁴³ International law may lack a rule of recognition on which the “validity thesis” rests. Still, a good approximation would be that one recognizes as law “that to which States have consented”,⁴⁴ even if one can doubt the centrality of consent in current international law.⁴⁵ The criteria for rules to be “presumptively sufficient” and seen as valid can be specified, for instance, by reference to the processes for States to express consent to be bound. The law of treaties has several examples, including the rules on authority to represent States and the procedures to expressly consent to be bound.⁴⁶

The “validity thesis” is not an aspect separate from the truth of the premises of legal syllogisms. The truth of those premises bleeds into the idea of “validity”. A “valid” rule, based on which one can construct a conclusive legal syllogism, must also be true. “Italy's territorial sea can extend up to 12 nm from its coast” cannot result from a conclusive syllogism unless it is true that, when one constructs the syllogism, the maximum breadth of the territorial sea under international law is 12 nm. If “validity” did not include that the relevant rule is a true reflection of the law, a legal syllogism could be formally valid but untrue. At a minimum, the purpose of the “validity thesis” and the conception of “validity” should also require rules to be true, in the sense of being an accurate reflection of the rules created by the accepted law-making mechanisms within the relevant legal system at a specific point in time.

C. Persuasiveness of Legal Reasoning in Custom Identification

Legal reasoning justifies judicial decisions concerning whether customary rules exist in custom identification. As the existence of customary rules depends on whether the elements of the definition of custom are satisfied, the purpose of legal reasoning is to provide good justifying reasons as to whether those elements are satisfied. It follows that one's definition of custom is the touchstone for assessing the persuasiveness of different forms of legal reasoning. According to the two-element theory, custom stems from the union of State practice and opinio juris. However, not all scholars accept this definition unreservedly. Some have proposed alternative narratives of custom.⁴⁷ Nevertheless, the two-element theory is firmly entrenched as the prevailing definition of custom, including in the ICJ's jurisprudence and the ILC's Draft Conclusions.⁴⁸ The question is, therefore, which form of reasoning can provide the best reasons justifying judicial decisions on the existence of customary international law defined by the two-element theory.

This question, in principle, requires a case-specific answer, but it appears possible to make some broad considerations. Legal reasoning is not simply persuasive or not. Persuasiveness is a spectrum along which different forms of reasoning position themselves. On the “higher persuasiveness” end of the spectrum, one finds inductive reasoning, which alone can show whether the elements of the definition of custom are satisfied. The premises of inductive reasoning do not compel their conclusions but have a foothold in empirical observation. Analysing evidence of State practice and opinio juris does not compel conclusions about whether customary rules exist, but one can support those conclusions by reference to empirical evidence of the constituent elements of custom. The premises of perfect syllogisms compel certain conclusions, thus perfect syllogisms situate themselves on the “higher persuasiveness” end of the spectrum too. The question is whether it is possible to construct perfect syllogisms to identify customary international law, which is discussed below.⁴⁹

Reasoning by analogy does not seek to show that the elements of the definition of custom are satisfied. Its premises do not compel certain conclusions either. Reasoning by analogy thus falls within the “lower persuasiveness” end of the spectrum. By definition, imperfect syllogisms do not compel conclusions unless combined with other forms of reasoning. For the purposes of this article, it is unnecessary to be prescriptive as to whether analogical reasoning is more persuasive than an imperfect syllogism, or the other way around.

Forms of reasoning can be combined in custom identification. The persuasiveness of perfect syllogisms is not increased by combining them with other forms of reasoning because they are already entirely persuasive. The combination of analogical reasoning with imperfect syllogisms remains in the “lower persuasiveness” end of the spectrum as neither form of reasoning can show that the elements of the definition of custom are satisfied in any given case. At best, combining analogical reasoning with imperfect syllogisms can be more persuasive than either form of reasoning used individually. When used alongside reasoning by analogy or imperfect syllogisms, inductive reasoning might appear more persuasive when used alone. This view would be misleading. When inductive reasoning is combined with other forms of reasoning, the ICJ has less incentive to review evidence of State practice and opinio juris thoroughly. While such a review is necessary to establish persuasively that the elements of the definition of custom are satisfied, the Court may emphasise one element over the other⁵⁰ or even downplay both.⁵¹ Instead, identifying custom by purely inductive reasoning pushes the Court to review more evidence of State practice and opinio juris than it otherwise would, as it did in Jurisdictional Immunities of the State. In practice, the combination of inductive reasoning and other forms of reasoning waters down the persuasiveness of the analysis into the existence of the elements of the definition of custom. For this reason, that combination is less persuasive than purely inductive reasoning.

The spectrum of persuasiveness described above remains the same regardless of whether the customary rules to be identified are prescriptive, permissive or prohibitive. For practice to support the existence of prescriptive or permissive rules, States must do the required or permitted conduct. The presence of a body of practice makes it feasible for the Court to identify prescriptive and permissive rules by inductive reasoning. Conversely, it might appear that the Court cannot persuasively identify prohibitive rules by inductive reasoning because State practice supporting the existence of such rules consists of omissions. However, State practice as omissions is still State practice.⁵² It might be unclear whether omissions are motivated by compliance with a prohibitive rule, which would support the existence of that rule, or by other reasons. Evidence of opinio juris clarifies the motives of omissions, which determines whether omissions support the existence of prohibitive rules.⁵³ In some cases, there could be no evidence of opinio juris to shed light on the motives of omissions by States. In such cases, not only would it be unclear whether State practice supports the existence of customary rules, but the absence of opinio juris would, in and of itself, mean that the Court could not identify the relevant customary rules.

II. Applying Deductive Reasoning to Custom Identification

Understood as above, pure deduction does not provide good justifying reasons for decisions to identify customary international law. The inadequacy of pure deduction as a form of reasoning for cutstom identification is discussed for three types of deduction: normative, functional and analogical.⁵⁴

A. Normative Deduction

In normative deduction, one infers new rules of international law from established ones. The ICJ often formulates such inferences by stating that new rules are corollaries of established ones. That a rule is the corollary of another can mean either that a new rule is part of an established one or that an established rule implies the existence of a new one as, without the new rule, the established one loses all or part of its significance. Cases like Jurisdictional Immunities of the State and Certain Documents and Data,⁵⁵ where new rules are said to “derive” from established ones, are comparable to the latter understanding because this process of “deriving” new rules from established ones appears to be based on a relationship of implication.

In Nicaragua, the Court considered whether the rule on non-intervention had customary status: first, it found that there was evidence of opinio juris backed by State practice; second, it found that non-intervention was a “corollary of the principle of the sovereign equality of States”.⁵⁶ This second stage should be capable of expression as a legal syllogism. One should work out the premises from the conclusion of this supposed syllogism, that “the rule on non-intervention is part of customary international law”, and from the basis for the inference, namely the principle of the sovereign equality of States. On the view that a new rule being the “corollary” of an established one means that the former is part of the latter, one could express the Court's statement as follows:

The principle of sovereign equality of States is part of customary international law;

AND

The rule on non-intervention is part of the principle of sovereign equality of States;

THEREFORE

The rule on non-intervention is part of customary international law.

This syllogism is conclusive, but the question is whether it is true. The truth of the major premise is beyond serious doubt, but one cannot necessarily say the same of the minor premise. Intuitively, if States are sovereign and equal, no State can intervene in the affairs of another. For the minor premise to be true, which is required in a legal syllogism, the content of the principle of sovereign equality must include the rule on non-intervention. Doubts can arise in this regard for two reasons. First, the principle of sovereign equality is indeterminate: it is so fundamental to the international legal system that one may trace virtually all rules of international law back to it.⁵⁷ One may avoid such indeterminacy if, before using deduction to infer the existence of a new customary rule, one clarifies the content of the established customary rule from which to infer it. This exercise would require ascertaining the existence of the established customary rule at a given point in time based on evidence of State practice and opinio juris.⁵⁸ By determining the content of the principle of sovereign equality of States based on State practice and opinio juris, one would aim to understand whether the rule on non-intervention is part of it. In this exercise, the evidence of relevant State practice and opinio juris would have to relate to non-intervention. This process would thus result in ascertaining whether the rule on non-intervention exists as a separate, new customary rule.

Second, the ICJ sought to ascertain the existence of a rule on non-intervention as a rule separate from the principle of sovereign equality. However, it is unusual for a rule, which is assumed to exist separately from another rule or principle, to be seen as part of that other rule or principle. It might be for this reason that the Court had the intuition, or calculated prudence, to state that the rule on non-intervention was a “corollary” of the principle of sovereign equality, meaning that it followed from it, rather than stating that it was “part of” that principle. The syllogism above would only have worked if the Court had stated that the rule on non-intervention was part of the principle of sovereign equality, which it did not. If the Court had so stated, the rule on non-intervention and the principle of sovereign equality could have been seen as the same rule, inconsistent with the Court's task of ascertaining the existence of non-intervention as a separate customary rule.

In another view, a new rule being the “corollary” of an established one means that the latter implies the existence of the former because, without the new rule, the established one can lose significance. On that view, the Court's statement in Nicaragua can be expressed as follows:

The principle of sovereign equality of States is part of customary international law;

AND

Without the rule on non-intervention, the principle of sovereign equality of States loses all or part of its significance;

THEREFORE

The rule on non-intervention is part of customary international law.

These two premises do not compel the conclusion. The only conclusion that can follow from these premises is that the rule on non-intervention must be part of international law. Still, nothing in this syllogism compels that rule to be part of customary international law. The principle of sovereign equality would not lose significance if the rule on non-intervention only existed as treaty law. Several multilateral treaties include a rule on non-intervention.⁵⁹ The case could be where the relevant rule does not exist as treaty law, which might require it to exist in customary law. The inexistence of the relevant rule as treaty law would be information additional and external to the premises of a syllogism of the type formulated above. This syllogism would be imperfect and thus inconclusive. The problems of framing normative deduction by syllogism will likely arise in other cases, such as the ICJ's statement in the Wall Opinion that the illegality of territorial acquisition resulting from the threat or use of force is a customary rule because it is a corollary of the customary prohibition on the threat or use of force.⁶⁰

B. Functional Deduction

Functional deduction entails inferring new rules of international law from the functions of persons or organizations. In its 2002 judgment in Arrest Warrant, the ICJ held that there was a customary rule pursuant to which “the functions of a Minister for Foreign Affairs are such that, throughout the duration of his or her office, he or she when abroad enjoys full immunity from criminal jurisdiction and inviolability”.⁶¹ The elements to build a syllogism to express the Court's reasoning are that Foreign Ministers enjoy immunity under customary international law and that, in customary international law, Foreign Minister immunity aims to ensure the effective performance of their functions.⁶² One can express the Court's reasoning by this syllogism:

Foreign Ministers enjoy immunity by virtue of customary international law;

AND

The immunity of Foreign Ministers aims to ensure the effective performance of their functions on behalf of their respective States;

THEREFORE

Under customary international law, Foreign Ministers enjoy full immunity from the criminal jurisdiction of foreign States and inviolability.

There seems to be no question as to the truth of the major and minor premise. One problem is that, in the minor premise, it is unclear what “effective performance” means. It is unclear whether to be “effective” means that a Foreign Minister's performance must be wholly unhindered by the exercise of a foreign State's jurisdiction or whether certain exercises of jurisdiction can be tolerated as not limiting the effectiveness of that performance.

Lack of clarity impacts the conclusion of this syllogism, which is a leap from its premises. That a Foreign Minister enjoys complete immunity from foreign criminal jurisdiction is a way to ensure their performance is effective, but it does not seem to be the only one. It might not hinder the effectiveness of a Foreign Minister's performance if they are prosecuted for crimes committed abroad that are not punished with custodial sentences. The syllogism expressing the Court's reasoning in Arrest Warrant thus emerges as imperfect. Other cases where the ICJ can infer the existence of customary rules from the functions of persons or organizations conceivably suffer from the same problems of indeterminacy arising in Arrest Warrant.⁶³ Indeterminacy problems may be avoided if such functions were exhaustively and clearly defined, which seems, if not impossible, at least unlikely because of the vagueness typical of legal norms. This vagueness alone does not mean that functional deduction is always unhelpful, but that the usefulness of functional deduction is limited to a few cases where the functions of persons or organizations from which to infer customary rules are part of the “core of settled meaning” and not of the “penumbra of uncertainty”.⁶⁴

In Arrest Warrant, Judge ad hoc Van den Wyngaert criticized the Court for assimilating the functions of Foreign Ministers to those of Heads of State, whose complete immunity from foreign criminal jurisdiction is not as controversial as that of Foreign Ministers.⁶⁵ Her criticism suggests another way to formulate a syllogism to express the Court's reasoning. This other formulation builds on the comparison between the functions of a Head of State and the functions of a Foreign Minister who, according to the Court, “like the Head of State […] is recognized under international law as representative of the State solely by virtue of his or her office”.⁶⁶ One could formulate this syllogism as follows:

In customary international law, the functions of Heads of State justify their entitlement to complete immunity from foreign criminal jurisdiction;

AND

The functions of Heads of State are the same as those of Foreign Ministers;

THEREFORE

In customary international law, the functions of Foreign Ministers justify their entitlement to full immunity from foreign criminal jurisdiction.

This syllogism is formally valid but also problematic. Even accepting the major premise as true, which it may not be as Heads of State can be held accountable for crimes under international law,⁶⁷ the minor premise is likely untrue. This syllogism builds on the unverified view that the Head of State and Foreign Minister functions are identical. As Judge ad hoc Van den Wyngaert wrote, this view may be an inaccurate reflection of the law, which raises issues under the “validity thesis”.⁶⁸ It may not be a valid rule in international law that Head of State and Foreign Minister functions are identical, which prevents this syllogism from being conclusive.

Furthermore, functional deduction builds on the assumption that one can infer the existence of customary rules from the functions of persons or organizations. This inference may appear intuitively possible where, as with State immunity, one justifies the existence of customary rules based on the need to protect the exercise of certain functions. However, the link between this need and the character of certain rules as custom is far from clear. This link would be between, on the one hand, the rationale of a rule whose existence is disputed and, on the other hand, the formal source of that rule. There is no logical reason why the character of a rule as customary would flow from that rule's rationale. As in normative deduction, this implication would be unsubstantiated by evidence of State conduct but would be assumed because of the absence of an applicable treaty rule of equivalent content to a supposed customary rule. This assumption makes functional deduction syllogisms imperfect and, thus, inconclusive, including where the functions of persons or organizations from which to infer the existence of customary rules are part of the “core of settled meaning” and not of the “penumbra of uncertainty”.⁶⁹

C. Analogical Deduction

By analogical deduction, one extends the rationale of existing rules to cases not falling within the wording of those rules because of a common cause or link between two situations. In delimiting the continental shelf between Libya and Malta in 1985, the ICJ assessed the relationship between the continental shelf and Exclusive Economic Zone (EEZ), codified three years earlier in the 1982 United Nations Convention on the Law of the Sea (UNCLOS).⁷⁰ The Court concluded that “the distance criterion must now apply to the continental shelf as well as to the [EEZ]”.⁷¹ One can express the ICJ's reasoning with this syllogism:

The distance criterion defines the geographical extent of a State's rights in the EEZ;

AND

A State's rights in the seabed and subsoil of the EEZ are defined by reference to its rights in the continental shelf;

THEREFORE

The distance criterion applies to the continental shelf.

The major premise is true. The Court had stated, before drawing its conclusion in relation to the continental shelf, that a customary rule existed under which “the institution of the [EEZ], with its rule on entitlement by reason of distance, is shown by the practice of States to have become part of customary law”.⁷² The minor premise also seems true, the only caveat being that the legal basis for the Court to find that EEZ rights are defined by reference to continental shelf rights likely was Article 56(3) of UNCLOS, which was not in force at that time.⁷³ The issue with this syllogism is that its premises do not compel its conclusion. If EEZ rights are defined by reference to the continental shelf regime, the criterion of continental shelf entitlement should apply to the EEZ, not the other way around. The Court had also remarked that “the institutions of the continental shelf and the [EEZ] are different and distinct”,⁷⁴ which casts additional doubts on whether the criterion of entitlement by reason of distance could be extended from the latter to the former.

Moreover, analogical deduction conflates two different types of legal reasoning: reasoning by deduction and reasoning by analogy. While the conclusions of legal syllogisms are compelled by their premises, analogical reasoning never compels a conclusion.⁷⁵ The ICJ's reasoning in Libya/Malta can support its decision that the distance criterion applies to the continental shelf, as the statement of principle that the distance criterion applies to the EEZ has a high degree of authority. There is an analogy, although not perfect, between the regime of the continental shelf and the regime of the EEZ. Nevertheless, the ICJ's reasoning does not compel that decision. In principle, it seems possible for a perfect analogy to compel a certain conclusion. But a perfect analogy can only exist where the unregulated situation to which one wishes to extend an established statement of principle already falls within the scope of that statement. There would be a perfect analogy if the EEZ and continental shelf regimes overlapped. In that case, the criterion of continental shelf entitlement would already be identical to the criterion of EEZ entitlement.

When relying on reasoning by analogy to identify customary rules, judges exercise a significant measure of discretion in appreciating whether the similarities between the situations to be compared carry more weight than their differences. In Libya/Malta, the Court acknowledged that natural prolongation remained a criterion of continental shelf entitlement, but found it to be “in part defined by distance from the shore” where the continental shelf extends up to 200 nm from the coast.⁷⁶ The Court could have equally decided that natural prolongation was a criterion of entitlement separate to and uninfluenced by distance from the coast, giving more weight to the difference between the continental shelf and the EEZ. Instead, the Court gave more weight to a similarity. The problem with such discretionary appreciations is that, on their basis, judges are rarely able persuasively to justify their decisions that certain customary rules exist. Writing on analogical reasoning in relation to common law precedent, legal theorists have endeavoured to formulate a framework to distinguish “material facts”, which are those facts that alone can justify analogies between cases, from other facts that are not relevant in establishing analogies.⁷⁷ However, there is no hard-and-fast rule in that regard.⁷⁸ A wide margin of discretion appears inevitable and, in ascertaining whether binding customary rules exist, it is hardly desirable because it would adversely affect legal certainty.

Reasoning by analogy can seem to resemble deductive reasoning. However, it appears more similar to inductive reasoning on account of its reliance on the historical evolution of rules or precedents.⁷⁹ Identifying customary rules by analogy is operationally similar to identifying those rules based on evidence of State practice and opinio juris, which requires assessing the evolution of State conduct over time. Because of their conceptual differences, it seems justified to separate legal syllogisms from analogical reasoning. At most, one can compare analogical reasoning to imperfect syllogisms as neither can compel their conclusions unless additional elements are introduced that are external to the analogy or the syllogism.

III. Authority, Utility, and Content Determination

The inadequacy of pure deduction as a form of reasoning to identify customary international law has some implications: first, one may question the authority of customary rules identified by pure deduction; second, one queries whether the Court could derive any utility from identifying custom by pure deduction; third, identifying customary rules by deduction could raise issues as to the determination of their content.

A. Authority of Custom Identification by Deduction

Purely deductive reasoning fails to provide good justifying reasons for identifying customary rules. The question is whether and how this failure affects the authority of decisions that certain rules are customary in character and, as a reflection, the authority of those very rules.

1. Low authority of custom identification by deduction

The authority of judicial decisions is linked, and even overlaps, with their legitimacy.⁸⁰ Broadly, legitimacy is the right to rule.⁸¹ In relation to rules, legitimacy is the belief that a rule ought to be obeyed.⁸² This belief exerts a pull towards compliance, which stems also from the perception that a rule has come into existence pursuant to accepted law-making processes.⁸³

Scholars have distinguished different facets of legitimacy. A first distinction is between normative legitimacy, the right to rule according to predefined standards, and sociological legitimacy, which is the perception within the relevant constituency that an institution has the right to rule.⁸⁴ Another distinction is between input legitimacy, linked to whether decisions are made pursuant to established procedures,⁸⁵ and output legitimacy, which overlaps with effectiveness and depends on whether decisions address the concerns of their constituency.⁸⁶ These distinctions are not clear-cut, but one may consider that the authority of decisions identifying customary rules chiefly concerns sociological and output legitimacy. A proper inquiry into sociological legitimacy would require an empirical study of attitudes within the relevant constituency, primarily made of States, concerning the ICJ's decisions identifying customary international law. Such a study is beyond the scope of this article. Judicial reasoning is also connected to the output legitimacy of court decisions.⁸⁷ The perceived quality of the ICJ's reasoning can impact whether its decisions achieve the concerns of States, which extend to knowing whether particular rules of international law exist that ought to guide their conduct.

It is by assessing the persuasiveness of legal reasoning that one should appreciate the authority of the ICJ's decisions that certain rules are part of customary international law. If the Court makes such decisions purely on the basis of deductive reasoning, their persuasiveness is in doubt.⁸⁸ One might even be justified in considering that customary rules identified by pure deduction have no persuasive basis in legal reasoning, which would compromise the authority of decisions identifying them and warrant the view that such rules are not rules at all. In such a scenario, legal reasoning would play, at most, a modest role in endowing custom identification with authority. The situation would be different where deductive reasoning is an aid to custom identification by induction, as in the Chagos opinion. Imperfect syllogisms have a foothold in empirical analysis so long as they are used to support custom identification alongside inductive reasoning. Evidence of State practice and opinio juris can enhance the authority of decisions on custom identification, even when it is not listed in detail in the relevant judicial decisions. However, in this scenario the authority-enhancing function of inductive reasoning appears more limited than it could otherwise be.⁸⁹

2. Sources of authority other than legal reasoning

One may have to temper the view that customary rules identified by pure deduction have weak authority. Judicial decisions can derive authority from sources other than the reasoning recorded in them. For customary rules identified by pure deduction, there seem to be at least three such sources that may confer legitimacy on decisions on custom identification.

First, legitimacy could flow from the implementation by the ICJ of established decision-making procedures, which include public adversarial proceedings and composite deliberation processes. The ICJ's pronouncements that customary rules exist are also decisions of the “principal judicial organ of the United Nations”,⁹⁰ which confers authority on those pronouncements because of the Court's legitimacy capital accumulated over more than seventy years of continuous operation.

Second, legitimacy could flow from the perception that customary rules identified by deduction may have a morally compelling content.⁹¹ Scholars have argued that deduction can be an appropriate form of reasoning to identify certain types of customary rules. Among such rules would be those flowing from the “constitutional foundations of the international community”⁹² or those seeking to promote moral values that State practice seems to contradict.⁹³ An example of the latter would be the customary rule on non-intervention, as identified by the ICJ in Nicaragua.⁹⁴ These views appear problematic because they leave open questions concerning morality in law-making, which legal scholars have been debating for decades without resolution.⁹⁵ Such questions concern not only which customary rules promote sufficiently important moral values to justify their identification by deductive reasoning, but also whose morality those rules should promote. The answer to the former question could be that, although it is difficult to agree on all rules that have a sufficiently important moral content, it may be possible to agree on a core of such rules.⁹⁶ The latter question is difficult to answer without falling back on moral relativism. That question relates to the role of morality which, conveyed by the idea of the “sacred trust of civilization”, was used also to justify the European colonial enterprise.⁹⁷ That question relates to the tension between the values promoted by different actors in international law as well. On one hand, States likely promote values that preserve their sovereign prerogatives. On the other hand, non-governmental actors, including scholars, are more likely to promote values limiting those sovereign prerogatives. Arrest Warrant exemplifies this tension. The Court's finding that Foreign Ministers are completely immune from foreign criminal jurisdiction under customary international law seems inconsistent with the morally compelling need to prosecute them for allegedly committing war crimes and crimes against humanity.⁹⁸ However, that finding is consistent with preserving a State's sovereign prerogatives, in this case the efficient discharge of ministerial functions. Arrest Warrant also shows that customary rules identified by deductive reasoning can be inconsistent with the moral values which, according to some, international law promotes or should promote. Identification by deductive reasoning does not necessarily result in ascertaining the existence of rules promoting moral values, which State practice contradicts.

Third, authority can derive from the likelihood that custom identification by deductive reasoning may reflect the changed structure of contemporary international law and, in this sense, respond to the relevant constituency's concerns. Inductive reasoning rooted in evidence of State behaviour is underpinned by the aim of protecting State sovereignty and thus fits a State-centric conception of international law. Deductive reasoning builds on the view that international law has become, to an appreciable extent, a semi-constitutionalized structure in which States come together in multilateral fora to address common challenges.⁹⁹ In this structure, rules derive authority from hierarchically superior ones.¹⁰⁰ This semi-constitutionalized structure sits uneasily with a State-centric conception of international law that views States as independent agents free from the commands of higher authorities.¹⁰¹ Defining the relevant constituency is crucial to assess whether deductive reasoning reflects the structure of contemporary international law. The relevant constituency certainly includes States. The question is whether it also includes non-State actors. If not, inductive reasoning would accurately reflect its makeup, which would mean that custom identification by induction would have greater authority than custom identification by deduction. If yes, one might be justified in deriving lower-level rules from higher-level ones by deductive reasoning because, beyond the former deriving their authority from the latter, both sets of rules would aim to address challenges common to all members of the constituency.

It is difficult to take a position either way. While an entirely State-centric view of international law is outdated, it remains difficult to explain away the continued centrality of States in customary international law.¹⁰² One can find suggestions in either direction in recent works purporting to codify certain rules of international law. The ILC's 2018 Draft Conclusions take an eminently State-centric view, in which States are the makers of custom and the source of all evidence of the elements of custom. Conversely, the 2005 Study of the Red Cross on Customary International Humanitarian Law often identifies customary rules by non-inductive processes, as do both editions of the Tallinn Manual on international law applicable to cyber warfare.¹⁰³ The approaches taken by these works are hardly surprising. The ILC is a body of experts elected by States operating within an international organization of which only States can be members. Because the Red Cross aims to ensure humanitarian protection in armed conflict, its 2005 Study was intended to ascertain the existence of as many rules limiting States in armed conflict as possible. The Tallinn Manual aimed to set out how rules of international law apply to cyberspace, but because there is still limited State practice in relation to cyberspace there would have been no point in writing the Manual unless one accepted that non-inductive processes could be used to identify customary rules. It is difficult to say that the structure of international law has changed to the extent that identifying customary rules by deductive reasoning reflects that changed structure. To the contrary, the continued centrality of States, although perhaps reduced, indicates that using inductive reasoning remains a major source of authority for decisions identifying customary international law.

B. Utility of Custom Identification by Deduction

Purely deductive reasoning does not provide good justifying reasons to identify rules of customary international law and weakens the authority of the customary rules thus identified. However, the Court has continued to use deductive reasoning to support custom identification. The question is whehter there are good reasons for doing so.

Certain writers have suggested that one can explain resorting to deductive reasoning based on the unfeasibility of identifying certain customary rules by induction. There seem to be four scenarios in which inductive reasoning is impossible to use: first, where State practice does not exist because a question is too new; second, where State practice is inconclusive because it is conflicting or too disparate; third, where one cannot establish the opinio juris of States; fourth, where there is a discrepancy between the conclusions indicated by evidence of State practice and the conclusions indicated by evidence of opinio juris.¹⁰⁴ The Court may use deductive reasoning to identify customary international law where the relevant rules are so established that one does not need to use evidence of State practice and opinio juris to justify their existence.¹⁰⁵ The Court may also use deductive reasoning depending on whether the supposed customary rules are “facilitative” or “moral”. “Facilitative” rules promote co-operation and co-existence among States, but do not deal with moral issues. For this reason, State practice is prominent in establishing their existence. State practice is less central to establishing the existence of “moral” rules because such rules prescribe ideal standards.¹⁰⁶ This view provides a possible theoretical explanation of why the Court may use deductive reasoning to identify customary international law. However, that view does not seem to grapple with the underlying issue, which concerns whether sufficient evidence of State practice and opinio juris exists to satisfy the standard to conclude that customary rules exist.

At the same time, the use of deductive reasoning in custom identification is more than a question of not being able to meet a standard, whether that standard is the “extensive and virtually uniform” one in North Sea Continental Shelf ¹⁰⁷ or the “widespread and representative” one in the ILC's Draft Conclusion 8.¹⁰⁸ Whether there is enough evidence to meet the required standard could become a sticking point in the drafting of the Court's decisions, especially where the evidence leads some judges, but not others, to conclude that customary rules exist. In such decisions, the impetus to maximise votes in favour leads to avoiding justifications based on evidence of State practice and opinio juris, on which judges easily disagree, and to record custom identification by deductive reasoning, which may be acceptable to a wider majority.¹⁰⁹ The use of deduction appears likely where judges may agree on the result of a custom identification exercise but disagree on how to get to that result. Judges who may prefer to justify custom identification by inductive reasoning will likely not object to deductive justification as long as the outcome of the custom identification exercise is the same.¹¹⁰

The situations where the Court identifies custom by deduction have one aspect in common: the Court is under pressure to use deduction to identify customary rules because evidence of State practice and opinio juris is so inconclusive or so clearly conclusive that using inductive reasoning would risk splitting the Court down the middle or be unnecessary. Using deduction as a way to muster agreement in deliberations is problematic. Where there is insufficient evidence of State practice and opinio juris, a rigorous approach would be for the Court to conclude that the relevant customary rule does not exist. The availability of deductive reasoning to justify custom identification allows the Court to avoid finding that no customary international law governs a certain question.¹¹¹ The Court knows that, by using deductive reasoning, it would find that customart international law exists, which would pre-empt the conclusion that its reasoning should set out to reach. The Court would work back from its desired conclusion to frame its reasoning based on the approach that is more likely to support that conclusion, instead of consistently applying a standard approach to reach whichever conclusion, whether desirable or not, that approach may yield.

An additional reason for the Court to identify customary rules by deduction relates to the parties’ submissions. The Court does not actively seek out evidence that the parties did not put before it. The Court thus resembles a court of common law, which implements an adversarial system, rather than a court of civil law, the powers of which follow a more inquisitorial model.¹¹² If the parties have not provided evidence of State practice and opinio juris, the Court is unlikely to seek that evidence out for itself. Even in cases where only one of the parties does not provide such evidence, the Court may still identify custom by deduction. Arrest Warrant exemplifies this scenario. While the Democratic Republic of the Congo did not provide evidence of State practice or opinio juris supporting the existence of a customary rule on absolute Foreign Minister immunity,¹¹³ Belgium submitted practice in the form of participation in international agreements, national legislation and national court decisions.¹¹⁴ The Court referred to no evidence of the constituent elements of custom in its judgment.¹¹⁵ Jurisdictional Immunities of the State is the exception to the ICJ's practice of relying only on evidence submitted by the parties. In that case, the Court's judgment referred to a piece of evidence not submitted by either Germany or Italy. However, both parties had already provided ample evidence of State practice and opinio juris on which to frame custom identification as an inductive exercise.¹¹⁶ The Court only supplemented the extensive evidence of State practice and opinio juris submitted by the parties.

This modus operandi can guide the parties' litigation strategy before the Court. Parties can submit evidence of State practice and opinio juris depending on how strong their argument on custom identification is. If Party A believes to have sufficient evidence to satisfy the standard to prove that a customary rule exists or not, it should run no risk using induction to frame the custom identification process. Arrest Warrant shows that adducing evidence of State practice and opinio juris is no guarantee that the Court would use that evidence to record custom identification as an inductive exercise. If Party A thinks the evidence to be thinner than desirable, it should frame custom identification using deductive reasoning to conceal the weakness of an inductive argument. Conversely, reliance on evidence of State practice and opinio juris is a formidable instrument for Party B to undermine Party A's claim that a customary rule exists. An indication in this sense comes from the opinions of the judges who, dissenting on points of custom identification, have invariably framed their views around the lack of sufficient evidence of State practice and opinio juris.¹¹⁷

C. Determining the Content of Customary Rules Identified by Deduction

Even if custom identification by deduction is problematic from the standpoint of legal reasoning, the ICJ will likely use deductive reasoning to identify customary rules in future cases. A question likely to arise in such cases concerns how one determines the content of the relevant customary rules for the purposes of litigation. This question concerns the process by which judges “find out” what customary rules mean, an example of such a process being interpretation.¹¹⁸

The ILC's 2018 Draft Conclusions state that identification is a process to ascertain customary rules and wholly determine their content.¹¹⁹ Most scholars adopt the same view, implicitly or explicitly.¹²⁰ Certain commentators have recently begun arguing that, while identification is a process to ascertain the existence of customary rules, one determines the content of those rules by way of a separate, interpretive process.¹²¹ These commentators have not fully explored the potential for determining the content of custom where one identifies customary rules by deductive reasoning. Their views are not entirely convincing where they build on the position that one identifies customary rules by reference to evidence of State practice and opinio juris.¹²² However, the case for the interpretability of custom might be viable where one has to determine the content of customary rules identified by deductive reasoning. While one determines the meaning of customary rules identified by inductive reasoning based on evidence of State practice and opinio juris, one is not constrained by such evidence in determining the content of customary rules identified by deductive reasoning.

In Arrest Warrant, the ICJ had to determine the extent of a Foreign Minister's immunity from foreign criminal jurisdiction. The Court found that a customary rule existed under which Foreign Ministers are entitled to such immunity because of a comparison between their functions and the functions of Heads of State.¹²³ In its judgment, the Court recorded no consideration of evidence of State practice or opinio juris. Consequently, one can conceive the Court's exercise of determining the extent of Foreign Minister immunity as an exercise in finding out the meaning of the customary rule on Foreign Minister immunity. The Court did not carry out this exercise as one separate from the process to ascertain the existence of the relevant customary rule on immunity. If the Court had done so, it might have avoided some of the criticism against its approach, such as Judge ad hoc Van den Wyngaert's statement that it should not have assimilated the functions of Foreign Ministers to those of Heads of State.¹²⁴

Arrest Warrant indicates that the Court's approach is to determine the content of customary rules identified by deductive reasoning chiefly by reference to the content pleaded by one of the parties before it. In Arrest Warrant, the Democratic Republic of the Congo argued that Foreign Ministers enjoy full immunity from foreign criminal jurisdiction,¹²⁵ which the Court accepted in its judgment.¹²⁶ The Court adopted a comparable approach in its 2014 order on provisional measures in Certain Documents and Data. In that case, the Court had to decide whether Timor-Leste's claimed right to confidentiality of communications with its legal counsel plausibly existed under international law. The Court found that Timor-Leste's claimed right was plausible because, consistent with deductive reasoning, it saw that right as one “derived from the principle of the sovereign equality of States”.¹²⁷ As to the content of that right, in its pleadings Timor-Leste had referred to common law and international jurisprudence upholding that right and to analogies with diplomatic documents.¹²⁸ Australia's arguments aimed to qualify Timor-Leste's claimed right as non-absolute.¹²⁹ Although Timor-Leste had not been entirely clear as to the contours of its claimed right, the Court relied on its pleadings to define that right's subject matter (protection of communications and confidentiality) and scope (any documents and data prepared by counsel for use in arbitration and negotiation).¹³⁰

The Court's approach in Arrest Warrant and Certain Documents and Data is problematic. Even if the ICJ's judgments only bind the parties¹³¹ and advisory opinions are not binding,¹³² the Court's decisions are regarded as authoritative statements of what international law is. Once the Court has stated that a given customary rule exists and has a certain content, one can consider that rule to be binding on all States. Relying on the parties’ pleadings to determine the content of customary rules identified by deduction raises two problems: first, it generalizes the parties’ determinations of the content of those rules; second, it does not anchor the content determination process in the practice of other States which will be bound by those rules.¹³³ To determine the content of customary rules identified by deduction, one should use other means unrelated to, or at least not over-reliant on, the parties’ pleadings. A full inquiry into these means is beyond the scope of this article, but it seems it would have to accept the problematic proposition that one can identify customary rules by deductive reasoning.

IV. Inadequacy of Pure Deduction in Custom Identification

Purely deductive reasoning is inadequate to frame the process to identify rules of customary international law. This view builds on the understanding of legal reasoning as a process to ground judicial decisions in public justification, which has the practical aim of persuasion. This view also builds on the understanding that classic deductive logic and legal reasoning share some overlap. Legal reasoning aims to achieve the highest possible certainty as to whether a legal system attaches certain consequences to the existence of certain facts. One can achieve that objective by framing legal reasoning as perfect syllogisms, which are deductive propositions the premises of which compel certain conclusions. Conversely, one could not achieve that objective by framing legal propositions as imperfect syllogisms, the premises of which cannot compel certain conclusions without help from other forms of reasoning.

To achieve the aim of legal reasoning, custom identification by deduction should be capable of being expressed as perfect syllogisms. Examples from the ICJ's cases show that one cannot express custom identification as perfect syllogisms. Whether one considers examples of normative, functional or analogical deduction, syllogisms are inconclusive because, for instance, they build on assumptions that are unverified. This view does not prevent using deductive reasoning as a form of justification additional to solid evidence of State practice and opinio juris, but shows the insufficiency of pure deduction as a form of reasoning to sustain conclusions that rules of customary international law exist.

The inadequacy of purely deductive reasoning carries certain implications. First, one can doubt the authority of decisions identifying customary international law by pure deduction and of the very rules thus identified. Second, there may well be reasons for the ICJ to continue identifying customary rules by deduction despite the problems of doing so, which one could link to the Court's deliberation process and the impact of the parties’ failure to submit evidence of State practice and opinio juris. Third, custom identification by pure deduction, despite being problematic, leaves open the question of determining the content of customary rules thus identified. So far, the Court has over-relied on the parties’ arguments as a means of content determination, but this approach appears misguided. Identifying customary international law by induction is surely more laborious than doing so by deduction and risks raising more controversy in judicial deliberations. So long as one adopts the two-element theory of custom, recording identification as a deductive exercise weakens the persuasiveness of the reasoning that ought to justify finding that rules of customary international law exist.